Max Payne 2: The Fall of Max Payne

2014-05-18 07:38:03 pm

Game Page: http://speeddemosarchive.com/MaxPayne2.html

Max Payne 2: The Fall of Max Payne (Any %) (Individual Level)

Verifier Responses

Quote:

Reject all on the grounds that this is a troll submission and significantly below the standards the runner has displayed with his other submissions.

Certain levels include deliberate time wasting e.g. pointlessly watching cutscenes.

We've discussed this in the verifier discussion thread and whether or not this is good enough reason to reject is unclear and rest with the admins. However, I'm not aware of any other troll submissions and so setting a precedent seems the right thing to do.

Quote:

Generally this category should't exist as it's not about completing the game as fast as possible but about manipulating the in-game timer.

One of these manipulations is slow-mo reloading as it stops the timer but the game keeps slowly running.
Because of this there are no downsides to always slow-mo reload even if it's awful for the viewer and at worst it's actually beneficial to constantly do so
as it's in level 3-4 where Vinnie moves during the slow-mo reload.
You can also reposition the camera during a slow-mo reload allowing better aiming.

Also if somebody dies the level time is reduced by 2 seconds.
The effect of this ability to reduce in-game time is best showcased in 2-1 where the runner managed an impressive record of -1.67 seconds.
Or should have if the timer was capable of going into the negatives but instead the time is rounded to zero.
Unfortunately the game decided to add 0.17 s to final time for no discernible reason.

As such it's an accept on all runs purely because they all are new wr's in manipulating the in-game timer.

Decision: Reject

Reason: See below.

https://queue.speeddemosarchive.com/queue/verificationfiles/1030/

This run will be available for a month. After that these link(s) will no longer work.

Thread title:

Game Page: http://speeddemosarchive.com/MaxPayne2.html

Max Payne 2: The Fall of Max Payne (Any %) (Individual Level)

[b]Verifier Responses[/b]

[quote]Reject all on the grounds that this is a troll submission and significantly below the standards the runner has displayed with his other submissions.

Certain levels include deliberate time wasting e.g. pointlessly watching cutscenes.

We've discussed this in the verifier discussion thread and whether or not this is good enough reason to reject is unclear and rest with the admins. However, I'm not aware of any other troll submissions and so setting a precedent seems the right thing to do.[/quote]

[quote]Generally this category should't exist as it's not about completing the game as fast as possible but about manipulating the in-game timer.

One of these manipulations is slow-mo reloading as it stops the timer but the game keeps slowly running.
Because of this there are no downsides to always slow-mo reload even if it's awful for the viewer and at worst it's actually beneficial to constantly do so
as it's in level 3-4 where Vinnie moves during the slow-mo reload.
You can also reposition the camera during a slow-mo reload allowing better aiming.

Also if somebody dies the level time is reduced by 2 seconds.
The effect of this ability to reduce in-game time is best showcased in 2-1 where the runner managed an impressive record of -1.67 seconds.
Or should have if the timer was capable of going into the negatives but instead the time is rounded to zero.
Unfortunately the game decided to add 0.17 s to final time for no discernible reason.

As such it's an accept on all runs purely because they all are new wr's in manipulating the in-game timer.[/quote]

[b]Decision:[/b] Reject

[b]Reason:[/b] See below.

https://queue.speeddemosarchive.com/queue/verificationfiles/1030/

This run will be available for a month. After that these link(s) will no longer work.

UraniumAnchor

2014-05-18 07:38:20 pm

Not a walrus

This run sets a strange precedent for a couple of reasons. Rather than reject the ILs that are unquestionably in need of improvement and accept the rest, I'm rejecting them all. Why? Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs. At least one of the levels has deliberate time wasting in the form of watching a cutscene that you're able to skip, even though it doesn't count against the timer you're still supposed to skip them unless there's a good reason not to. Several of the levels have visible mistakes and appeared to save time because of new routing rather than better execution. For reference, here's how these runs compare to the current runs on the site (though there are some improvements working their way through the queue that may or may not be better than these):

1-P | Old = 0:14.27 | New = 0:14.11 | Diff = 0.16
1-1 | Old = 1:12.32 | New = 0:59.99 | Diff = 12.33
1-2 | Old = 0:09.67 | New = 0:09.28 | Diff = 0.39
1-3 | Old = 0:39.33 | New = 0:39.01 | Diff = 0.32
1-4 | Old = 0:58.61 | New = 0:48.44 | Diff = 10.17
1-5 | Old = 0:14.19 | New = 0:13.37 | Diff = 0.82
1-6 | Old = 1:28.81 | New = 1:28.42 | Diff = 0.39
1-7 | Old = 0:48.76 | New = 0:48.59 | Diff = 0.17

2-P | Old = 1:29.93 | New = 1:28.87 | Diff = 1.06
2-1 | Old = 0:03.43 | New = 0:00.17 | Diff = 3.26
2-2 | Old = 0:18.26 | New = 0:17.80 | Diff = 0.46
2-3 | Old = 0:16.91 | New = 0:16.76 | Diff = 0.15
2-4 | Old = 0:22.91 | New = 0:21.98 | Diff = 0.93
2-5 | Old = 0:51.72 | New = 0:49.19 | Diff = 2.53
2-6 | Old = 0:29.30 | New = 0:28.73 | Diff = 0.57

3-P | Old = 0:50.98 | New = 0:50.35 | Diff = 0.63
3-1 | Old = 0:27.48 | New = 0:26.11 | Diff = 1.37
3-2 | Old = 0:13.45 | New = 0:12.44 | Diff = 1.01
3-3 | Old = 0:09.44 | New = 0:08.44 | Diff = 1.00
3-4 | Old = 3:18.27 | New = 3:03.07 | Diff = 15.20
3-5 | Old = 0:35.31 | New = 0:34.79 | Diff = 0.52
3-6 | Old = 0:52.94 | New = 0:51.71 | Diff = 1.23
3-7 | Old = 0:13.78 | New = 0:11.14 | Diff = 2.64
3-8 | Old = 1:08.55 | New = 1:03.96 | Diff = 4.59

Total | Old = 17:28.62 | New = 16:26.72 | Diff = 1:01.90

I realize this decision is likely to spawn a heated argument. And that's fine. Just keep it civil.

However, I'm going to say this one more time, and I don't want to hear any more arguing about it unless you have new information to add:

The game timer is accurate enough that this is a legitimate category.

BUT

That does NOT mean you can waste time deliberately just because it does not count against you.

Unless somebody can show something I (and the other staff) don't already know to convince me that the game timer is not usable, then this is how the ruling will stay. This means that:

1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.
2) It is fine that the timer does not count cutscenes, but you still need to skip the cutscenes! This has been the rule for quite some time and this is not new.
3) It is fine that killing enemies reverses the timer, even if that means in some levels it's "faster" to go out of your way to kill extra enemies. If there's levels where this significantly changes the route then you could make an argument that "no extra enemies" could be its own category for that level, but that's a stretch and out of scope for this thread. It wouldn't be the first time where certain ILs get an extra category but others don't, but the difference would have to be significant. I don't know if any of the levels in this game have infinitely respawning enemies or anything like that, but if any of them do that might be an exception (as in, don't do that).
4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.

Also, a recent submission by a different runner has improved the p3c8 time in this submission by approximately 14 in-game seconds.

And to the two runners who submitted this table, not that it's a secret who they are:

If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.

Without the insulting run comments.

This run sets a strange precedent for a couple of reasons. Rather than reject the ILs that are unquestionably in need of improvement and accept the rest, I'm rejecting them all. Why? Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs. At least one of the levels has deliberate time wasting in the form of watching a cutscene that you're able to skip, even though it doesn't count against the timer you're still supposed to skip them unless there's a good reason not to. Several of the levels have visible mistakes and appeared to save time because of new routing rather than better execution. For reference, here's how these runs compare to the current runs on the site (though there are some improvements working their way through the queue that may or may not be better than these):

[hidden]
1-P	|	Old = 0:14.27	| New = 0:14.11 | Diff = 0.16
1-1	|	Old = 1:12.32	| New = 0:59.99 | Diff = 12.33
1-2	|	Old = 0:09.67	| New = 0:09.28 | Diff = 0.39
1-3	|	Old = 0:39.33	| New = 0:39.01 | Diff = 0.32
1-4	|	Old = 0:58.61	| New = 0:48.44 | Diff = 10.17
1-5	|	Old = 0:14.19	| New = 0:13.37 | Diff = 0.82
1-6	|	Old = 1:28.81	| New = 1:28.42 | Diff = 0.39
1-7	|	Old = 0:48.76	| New = 0:48.59 | Diff = 0.17

2-P	|	Old = 1:29.93	| New = 1:28.87 | Diff = 1.06
2-1	|	Old = 0:03.43	| New = 0:00.17 | Diff = 3.26
2-2	|	Old = 0:18.26	| New = 0:17.80	| Diff = 0.46
2-3	|	Old = 0:16.91	| New = 0:16.76	| Diff = 0.15
2-4	|	Old = 0:22.91	| New = 0:21.98 | Diff = 0.93
2-5	|	Old = 0:51.72	| New = 0:49.19 | Diff = 2.53
2-6	|	Old = 0:29.30	| New = 0:28.73 | Diff = 0.57

3-P	|	Old = 0:50.98	| New = 0:50.35 | Diff = 0.63
3-1	|	Old = 0:27.48	| New = 0:26.11 | Diff = 1.37
3-2	|	Old = 0:13.45	| New = 0:12.44 | Diff = 1.01
3-3	|	Old = 0:09.44	| New = 0:08.44 | Diff = 1.00
3-4	|	Old = 3:18.27	| New = 3:03.07 | Diff = 15.20
3-5	|	Old = 0:35.31	| New = 0:34.79 | Diff = 0.52
3-6	|	Old = 0:52.94	| New = 0:51.71 | Diff = 1.23
3-7	|	Old = 0:13.78	| New = 0:11.14 | Diff = 2.64
3-8	|	Old = 1:08.55	| New = 1:03.96 | Diff = 4.59

Total	|	Old = 17:28.62	| New = 16:26.72 | Diff = 1:01.90
[/hidden]

I realize this decision is likely to spawn a heated argument. And that's fine. Just keep it civil.

However, I'm going to say this one more time, and I don't want to hear any more arguing about it unless you have [b]new[/b] information to add:

[b]The game timer is accurate enough that this is a legitimate category.

[i]BUT[/i]

That does NOT mean you can waste time deliberately just because it does not count against you.[/b]

Unless somebody can show something I (and the other staff) don't already know to convince me that the game timer is not usable, then this is how the ruling will stay. This means that:

1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.
2) It is fine that the timer does not count cutscenes, but you still need to skip the cutscenes! This has been the rule for quite some time and this is not new.
3) It is fine that killing enemies reverses the timer, even if that means in some levels it's "faster" to go out of your way to kill extra enemies. If there's levels where this significantly changes the route then you could make an argument that "no extra enemies" could be its own category for that level, but that's a stretch and out of scope for this thread. It wouldn't be the first time where certain ILs get an extra category but others don't, but the difference would have to be significant. I don't know if any of the levels in this game have infinitely respawning enemies or anything like that, but if any of them do that might be an exception (as in, don't do that).
4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.

Also, a recent submission by a different runner has improved the p3c8 time in this submission by approximately 14 in-game seconds.

And to the two runners who submitted this table, not that it's a secret who they are:

If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.

[b]Without[/b] the insulting run comments.

HoboWithaShotgun

2014-05-18 08:33:58 pm

Wake up and be awesome

Okay, so I know that I said I wouldn't be able to verify runs or do much here in the forums not to long ago, but this is interesting

Quote:

And to the two runners who submitted this table, not that it's a secret who they are:

Alright, I'm an idiot, who are they? If you can't answer do to privacy reasons, that's fine

UraniumAnchor

2014-05-18 08:43:31 pm

Not a walrus

The videos have complete StatIDs on them, including the runner names.

Guhbadoo

2014-05-18 08:47:25 pm

Quote:

3) It is fine that killing enemies reverses the timer, even if that means in some levels it's "faster" to go out of your way to kill extra enemies. If there's levels where this significantly changes the route then you could make an argument that "no extra enemies" could be its own category for that level, but that's a stretch and out of scope for this thread. It wouldn't be the first time where certain ILs get an extra category but others don't, but the difference would have to be significant. I don't know if any of the levels in this game have infinitely respawning enemies or anything like that, but if any of them do that might be an exception (as in, don't do that).
4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.

How is this in any way fine? Random timers are not accurate so if this is not worth invalidating the timer nothing is. IGT seems to have no connection to real time so there's no logical reason to use it. Even contemplating making extra categories simply because of a bad timer is absurd. Speed should be the only thing that matters. Not some arbitrary number.

[quote]3) It is fine that killing enemies reverses the timer, even if that means in some levels it's "faster" to go out of your way to kill extra enemies. If there's levels where this significantly changes the route then you could make an argument that "no extra enemies" could be its own category for that level, but that's a stretch and out of scope for this thread. It wouldn't be the first time where certain ILs get an extra category but others don't, but the difference would have to be significant. I don't know if any of the levels in this game have infinitely respawning enemies or anything like that, but if any of them do that might be an exception (as in, don't do that).
4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.[/quote]

How is this in any way fine?  Random timers are not accurate so if this is not worth invalidating the timer nothing is.  IGT seems to have no connection to real time so there's no logical reason to use it.  Even contemplating making extra categories simply because of a bad timer is absurd.  Speed should be the only thing that matters.  Not some arbitrary number.

HoboWithaShotgun

2014-05-18 08:48:25 pm

Wake up and be awesome

Oh for god sakes, I knew I was an idiot, but not that big of an idiot Angry

What your saying is that you rejected this run, that even though it was faster, it just wasn't up to proper SDA format and such. Just want to be clear on taht

tekkie

2014-05-18 08:59:56 pm

~

Quote from Guhbadoo:

How is this in any way fine? Random timers are not accurate so if this is not worth invalidating the timer nothing is. IGT seems to have no connection to real time so there's no logical reason to use it. Even contemplating making extra categories simply because of a bad timer is absurd. Speed should be the only thing that matters. Not some arbitrary number.

the strength of which some people cling to faulty in-game timers is often pretty incredible

HoboWithaShotgun

2014-05-18 09:16:26 pm

Wake up and be awesome

In game timers are so popular, certain runs have gone to extremes to abuse this. I saw a run where the ingame timer was in the negatives

UraniumAnchor

Edit history:

UraniumAnchor: 2014-05-18 09:53:26 pm

UraniumAnchor: 2014-05-18 09:52:13 pm

UraniumAnchor: 2014-05-18 09:51:36 pm

UraniumAnchor: 2014-05-18 09:50:47 pm

2014-05-18 09:49:36 pm

Not a walrus

Quote from Guhbadoo:

How is this in any way fine? Random timers are not accurate so if this is not worth invalidating the timer nothing is. IGT seems to have no connection to real time so there's no logical reason to use it. Even contemplating making extra categories simply because of a bad timer is absurd. Speed should be the only thing that matters. Not some arbitrary number.

If it were more than 0.17 seconds I might agree with you, but that's such a tiny margin that it's not worth worrying about. The implication that skill does not matter just because the timer is weird is specious at best.

Quote from HoboWithaShotgun:

Oh for god sakes, I knew I was an idiot, but not that big of an idiot Angry

What your saying is that you rejected this run, that even though it was faster, it just wasn't up to proper SDA format and such. Just want to be clear on taht

I'm rejecting it because some of the levels are badly executed (sometimes deliberately) even if the time is better, and neither I nor the verifiers feel like sorting through which ones to keep given the nature of the submission. There's quite a bit of discussion that I didn't include in the response post, but I feel like I summarized it pretty well.

Any more posts about the validity of the game timer is getting deleted. We've rehashed this same discussion far too many times and we're not having it again. Not in this thread, at least. It's very simple: If you don't like this category, don't watch or run it. It's not going away. There's nothing stopping people from doing RTA ILs if they want to.

[quote="Guhbadoo"]How is this in any way fine?  Random timers are not accurate so if this is not worth invalidating the timer nothing is.  IGT seems to have no connection to real time so there's no logical reason to use it.  Even contemplating making extra categories simply because of a bad timer is absurd.  Speed should be the only thing that matters.  Not some arbitrary number.[/quote]

If it were more than 0.17 seconds I might agree with you, but that's such a tiny margin that it's not worth worrying about. The implication that skill does not matter just because the timer is weird is specious at best.

[quote="HoboWithaShotgun"]Oh for god sakes, I knew I was an idiot, but not that big of an idiot >:(

What your saying is that you rejected this run, that even though it was faster, it just wasn't up to proper SDA format and such. Just want to be clear on taht[/quote]

I'm rejecting it because some of the levels are badly executed (sometimes deliberately) even if the time is better, and neither I nor the verifiers feel like sorting through which ones to keep given the nature of the submission. There's quite a bit of discussion that I didn't include in the response post, but I feel like I summarized it pretty well.

Any more posts about the validity of the game timer is getting deleted. We've rehashed this same discussion far too many times and we're not having it again. Not in this thread, at least. It's very simple: If you don't like this category, don't watch or run it. It's not going away. There's nothing stopping people from doing RTA ILs if they want to.

HoboWithaShotgun

2014-05-18 10:01:06 pm

Wake up and be awesome

Alright, these probably are my grounds to fight on, I'll leave this discussion

Kotti

Edit history:

Kotti: 2014-05-19 05:38:01 am

2014-05-19 05:28:10 am

Quote:

Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs

While the intention may not have been to give the site better runs, the goal was still to improve the overall quality by getting rid of bad runs. The current ILs make both Max Payne 2 runs and SDA look bad. It's also not making SDA look very good when the verification system in case of this game clearly does not work. Last time ILs were submitted both verifiers rejected them as the improvements were insignificant and the runs still weren't fully optimized. The runs still got accepted because they were faster. The runs in this submission are faster and one of the verifiers accepted them yet the decision is to reject them. Why are these runs verified in the first place?

Quote:

1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.

If there's any waiting involved at all, reloading saves IGT. This is most noticeable in 3-4 which is basically an autoscroller so every single slowmo reload saves a tiny bit of time. No matter how bad it looks it is an actual timesaver that saves about 10 in-game seconds while costing several minutes of real time. I could also argue that it can be used like pause buffering as you still have some control over the camera making aiming easier but that really is a minor difference.

Quote:

4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.

How is it acceptable for the timer to be a part of the RNG? In most cases I would agree that inaccuracy of 0.17 seconds is insignificant but here the entire run is nothing but those 0.17 seconds so it makes a huge difference. The best run is literally just the run where the timer chooses to be the least inaccurate.

Quote:

If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.

How is that defined? 3-1 for example has 30+ slowmo reloads in it but at the same time it is very optimized and beating it would be fairly challenging. Optimizing the time and doing a million slowmo reloads are not mutually exclusive actions in case of this category so a run that looks terrible can still be very good.

Quote:

Without the insulting run comments.

They may have been excessive but at least they were more up to date than the current comments that don't comment on the issues with the timer at all, have references to runs that are no longer on the site and have incorrect explanations for the tricks.

Quote:

Any more posts about the validity of the game timer is getting deleted

Best way to solve a problem is look away and pretend it's not there Wink

In my opinion there are two possible sensible solutions - Either admit that the ruleset used for this game is broken and easily exploitable and accept the runs or fix the ruleset.

[quote]Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs[/quote]While the intention may not have been to give the site better runs, the goal was still to improve the overall quality by getting rid of bad runs. The current ILs make both Max Payne 2 runs and SDA look bad. It's also not making SDA look very good when the verification system in case of this game clearly does not work. Last time ILs were submitted both verifiers rejected them as the improvements were insignificant and the runs still weren't fully optimized. The runs still got accepted because they were faster. The runs in this submission are faster and one of the verifiers accepted them yet the decision is to reject them. Why are these runs verified in the first place?
[quote]1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.[/quote]If there's any waiting involved at all, reloading saves IGT. This is most noticeable in 3-4 which is basically an autoscroller so every single slowmo reload saves a tiny bit of time. No matter how bad it looks it is an actual timesaver that saves about 10 in-game seconds while costing several minutes of real time. I could also argue that it can be used like pause buffering as you still have some control over the camera making aiming easier but that really is a minor difference.
[quote]4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.[/quote]How is it acceptable for the timer to be a part of the RNG? In most cases I would agree that inaccuracy of 0.17 seconds is insignificant but here the entire run is nothing but those 0.17 seconds so it makes a huge difference. The best run is literally just the run where the timer chooses to be the least inaccurate.
[quote]If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.[/quote]How is that defined? 3-1 for example has 30+ slowmo reloads in it but at the same time it is very optimized and beating it would be fairly challenging. Optimizing the time and doing a million slowmo reloads are not mutually exclusive actions in case of this category so a run that looks terrible can still be very good.
[quote]Without the insulting run comments.[/quote]They may have been excessive but at least they were more up to date than the current comments that don't comment on the issues with the timer at all, have references to runs that are no longer on the site and have incorrect explanations for the tricks.
[quote]Any more posts about the validity of the game timer is getting deleted[/quote]Best way to solve a problem is look away and pretend it's not there ;)

In my opinion there are two possible sensible solutions - Either admit that the ruleset used for this game is broken and easily exploitable and accept the runs or fix the ruleset.

Onin

2014-05-19 08:26:21 am

Honest question to UA:

I agree that even when the target of a run is the in-game timer, and there are various ways to manipulate it, RTA speed should still be kept in account as a secondary goal for the sake of entertainment/run quality. So no letting cutscenes play when they don't have to, no arbitrarily pausing the timer without gains. But in this case it's clear that a severe cost in RTA time can save some seconds of in-game time. Is there some sort of cut-off point, even if arbitrary, or would a runner be expected to abuse every trick for a lower in-game time no matter the cost in RTA?

The reason this question is important isn't just for this verification. There will be other games with weird in-game timers, where wasting real time in order to save in-game time will be a thing. Is it okay to waste a minute of real time for a 10 seconds faster in-game timer? What about 5 minutes? 10?

PJ

2014-05-19 10:09:18 am

Is PJ

I agree with Kotti's post. If the category is IGT, then manipulating IGT should be the main goal.

puwexil

2014-05-19 10:32:35 am

Professional Second Banana

Quote from UraniumAnchor:

There's nothing stopping people from doing RTA ILs if they want to.

This. We are perfectly happy with having separate game-time and real-time IL tables for this game given how different the play style is.

UraniumAnchor

Edit history:

UraniumAnchor: 2014-05-19 11:12:47 am

2014-05-19 11:12:09 am

Not a walrus

Quote from Kotti:

Quote:

Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs

While the intention may not have been to give the site better runs, the goal was still to improve the overall quality by getting rid of bad runs. The current ILs make both Max Payne 2 runs and SDA look bad. It's also not making SDA look very good when the verification system in case of this game clearly does not work. Last time ILs were submitted both verifiers rejected them as the improvements were insignificant and the runs still weren't fully optimized. The runs still got accepted because they were faster. The runs in this submission are faster and one of the verifiers accepted them yet the decision is to reject them. Why are these runs verified in the first place?

I'm still debating on whether or not I want to establish a minimum percentage improvement for these ILs, since "a second" is probably not enough for some of them, but for some of them a second is quite a lot. It's certainly raised a lot of questions that I'm not sure I have a good answer for.

ILs are easily the most work to keep updated, but at the same time I don't want to discourage significant improvements. The question here is what counts as significant. Maybe "two percent, or a second, whichever is less".

Quote from Kotti:

Quote:

1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.

If there's any waiting involved at all, reloading saves IGT. This is most noticeable in 3-4 which is basically an autoscroller so every single slowmo reload saves a tiny bit of time. No matter how bad it looks it is an actual timesaver that saves about 10 in-game seconds while costing several minutes of real time. I could also argue that it can be used like pause buffering as you still have some control over the camera making aiming easier but that really is a minor difference.

Right, and this counts as an actual good reason. The last bit of my statement is just there because I'm not sure if there's any points where you can spin reload purely to waste RTA time without getting some IGT gains back in return. Again, I haven't played this game so I have to go with what people are telling me and what I'm seeing in the videos.

Quote from Kotti:

Quote:

4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.

How is it acceptable for the timer to be a part of the RNG? In most cases I would agree that inaccuracy of 0.17 seconds is insignificant but here the entire run is nothing but those 0.17 seconds so it makes a huge difference. The best run is literally just the run where the timer chooses to be the least inaccurate.

I'm not entirely happy with it either, but I'm not about to throw out the whole table because the timer can be up to a sixth of a second off. It is unfortunate that the level in question is the one where it's the entire timer.

Quote from Kotti:

Quote:

If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.

How is that defined? 3-1 for example has 30+ slowmo reloads in it but at the same time it is very optimized and beating it would be fairly challenging. Optimizing the time and doing a million slowmo reloads are not mutually exclusive actions in case of this category so a run that looks terrible can still be very good.

Unfortunately I didn't get a great answer from the verifiers on this despite quite a bit of prodding. They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.

Quote from Kotti:

Quote:

Without the insulting run comments.

They may have been excessive but at least they were more up to date than the current comments that don't comment on the issues with the timer at all, have references to runs that are no longer on the site and have incorrect explanations for the tricks.

Right, but you can be up to date without being insulting. It comes across as petulant at best. I'm not the only one who thought they were unacceptable. If I had been, I wouldn't bring it up at all.

Quote from Kotti:

Quote:

Any more posts about the validity of the game timer is getting deleted

Best way to solve a problem is look away and pretend it's not there Wink

In my opinion there are two possible sensible solutions - Either admit that the ruleset used for this game is broken and easily exploitable and accept the runs or fix the ruleset.

It's more that unless there's new information to add I don't see the point in rehashing the argument, since there are several people (some runners, some staff, some both) who agree that the timer, while weird, is fine and it's just generating a lot of noise to argue about it. If there's something I don't know about the timer then that's different, but it seems like most of the posts about it are just people saying they disagree with the staff's interpretation.

Quote from PJ:

I agree with Kotti's post. If the category is IGT, then manipulating IGT should be the main goal.

I'm not sure when this was ever in debate, so I apologize if anything I said misled people. Is there something I should clarify?

[quote="Kotti"][quote]Primarily, because the submission was made in bad faith, intending to use it as some kind of argument against the existence of the category rather than to give the site better runs[/quote]While the intention may not have been to give the site better runs, the goal was still to improve the overall quality by getting rid of bad runs. The current ILs make both Max Payne 2 runs and SDA look bad. It's also not making SDA look very good when the verification system in case of this game clearly does not work. Last time ILs were submitted both verifiers rejected them as the improvements were insignificant and the runs still weren't fully optimized. The runs still got accepted because they were faster. The runs in this submission are faster and one of the verifiers accepted them yet the decision is to reject them. Why are these runs verified in the first place?[/quote]

I'm still debating on whether or not I want to establish a minimum percentage improvement for these ILs, since "a second" is probably not enough for some of them, but for some of them a second is quite a lot. It's certainly raised a lot of questions that I'm not sure I have a good answer for.

ILs are easily the most work to keep updated, but at the same time I don't want to discourage significant improvements. The question here is what counts as significant.  Maybe "two percent, or a second, whichever is less".

[quote="Kotti"][quote]1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.[/quote]If there's any waiting involved at all, reloading saves IGT. This is most noticeable in 3-4 which is basically an autoscroller so every single slowmo reload saves a tiny bit of time. No matter how bad it looks it is an actual timesaver that saves about 10 in-game seconds while costing several minutes of real time. I could also argue that it can be used like pause buffering as you still have some control over the camera making aiming easier but that really is a minor difference.[/quote]

Right, and this counts as an actual good reason. The last bit of my statement is just there because I'm not sure if there's any points where you can spin reload purely to waste RTA time without getting some IGT gains back in return. Again, I haven't played this game so I have to go with what people are telling me and what I'm seeing in the videos.

[quote="Kotti"][quote]4) It is fine that the timer adds some indeterminate/random amount of time at the end even after the timer appears to have stopped. The worst example I was shown was 0.17 seconds, so this is not worth invalidating the timer over. Consider it part of the RNG.[/quote]How is it acceptable for the timer to be a part of the RNG? In most cases I would agree that inaccuracy of 0.17 seconds is insignificant but here the entire run is nothing but those 0.17 seconds so it makes a huge difference. The best run is literally just the run where the timer chooses to be the least inaccurate.[/quote]

I'm not entirely happy with it either, but I'm not about to throw out the whole table because the timer can be up to a sixth of a second off. It is unfortunate that the level in question is the one where it's the entire timer.

[quote="Kotti"][quote]If you want to resubmit the levels that are actually played well (in terms of this category), you're welcome to do so.[/quote]How is that defined? 3-1 for example has 30+ slowmo reloads in it but at the same time it is very optimized and beating it would be fairly challenging. Optimizing the time and doing a million slowmo reloads are not mutually exclusive actions in case of this category so a run that looks terrible can still be very good.[/quote]

Unfortunately I didn't get a great answer from the verifiers on this despite quite a bit of prodding. They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.

[quote="Kotti"][quote]Without the insulting run comments.[/quote]They may have been excessive but at least they were more up to date than the current comments that don't comment on the issues with the timer at all, have references to runs that are no longer on the site and have incorrect explanations for the tricks.[/quote]

Right, but you can be up to date without being insulting. It comes across as petulant at best. I'm not the only one who thought they were unacceptable. If I had been, I wouldn't bring it up at all.

[quote="Kotti"][quote]Any more posts about the validity of the game timer is getting deleted[/quote]Best way to solve a problem is look away and pretend it's not there ;)

In my opinion there are two possible sensible solutions - Either admit that the ruleset used for this game is broken and easily exploitable and accept the runs or fix the ruleset.[/quote]

It's more that unless there's new information to add I don't see the point in rehashing the argument, since there are several people (some runners, some staff, some both) who agree that the timer, while weird, is fine and it's just generating a lot of noise to argue about it. If there's something I don't know about the timer then that's different, but it seems like most of the posts about it are just people saying they disagree with the staff's interpretation.

[quote="PJ"]I agree with Kotti's post.  If the category is IGT, then manipulating IGT should be the main goal.[/quote]

I'm not sure when this was ever in debate, so I apologize if anything I said misled people. Is there something I should clarify?

zastbat

Edit history:

zastbat: 2014-05-19 11:24:37 am

2014-05-19 11:23:41 am

Lord Of The Beards

Quote:

They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.

iirc watching certain cutscenes is better since for whatever reason enemies killed in them are counted towards lowering the IGT which would fall under manipulating IGT. although I have not watched all of the ILs so I have no idea if cutscenes that did not do this were left in. Someone can correct me if im wrong I might be remembering incorrectly.

UraniumAnchor

Edit history:

UraniumAnchor: 2014-05-19 11:26:27 am

2014-05-19 11:24:51 am

Not a walrus

Quote from Onin:

Honest question to UA:

I agree that even when the target of a run is the in-game timer, and there are various ways to manipulate it, RTA speed should still be kept in account as a secondary goal for the sake of entertainment/run quality. So no letting cutscenes play when they don't have to, no arbitrarily pausing the timer without gains. But in this case it's clear that a severe cost in RTA time can save some seconds of in-game time. Is there some sort of cut-off point, even if arbitrary, or would a runner be expected to abuse every trick for a lower in-game time no matter the cost in RTA?

The reason this question is important isn't just for this verification. There will be other games with weird in-game timers, where wasting real time in order to save in-game time will be a thing. Is it okay to waste a minute of real time for a 10 seconds faster in-game timer? What about 5 minutes? 10?

I don't really have a hard rule for this, but as an example of something we decided not to accept, one of the Sonic Heroes teams (Dark, I think?) has a move that freezes the game timer and lets you get 0:00.00 on essentially every level. It's more a question of just how much "skill" it removes from the result, not so much a % of RTA vs IGT. In this game, at least, you still have to execute well to get a good time, even if the resulting distinction looks tiny in some cases.

In general, though, yes, that's the whole reason we make the separation to begin with.

Quote from zastbat:

Quote:

They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.

iirc watching certain cutscenes is better since for whatever reason enemies killed in them are counted towards lowering the IGT which would fall under manipulating IGT. although I have not watched all of the ILs so I have no idea if cutscenes that did not do this were left in.

Right, I'm not referring to that, and neither was the verifier. This was a different cutscene.

[quote="Onin"]Honest question to UA:

I agree that even when the target of a run is the in-game timer, and there are various ways to manipulate it, RTA speed should still be kept in account as a secondary goal for the sake of entertainment/run quality. So no letting cutscenes play when they don't have to, no arbitrarily pausing the timer without gains. But in this case it's clear that a severe cost in RTA time can save some seconds of in-game time. Is there some sort of cut-off point, even if arbitrary, or would a runner be expected to abuse every trick for a lower in-game time no matter the cost in RTA?

The reason this question is important isn't just for this verification. There will be other games with weird in-game timers, where wasting real time in order to save in-game time will be a thing. Is it okay to waste a minute of real time for a 10 seconds faster in-game timer? What about 5 minutes? 10?[/quote]

I don't really have a hard rule for this, but as an example of something we decided not to accept, one of the Sonic Heroes teams (Dark, I think?) has a move that freezes the game timer and lets you get 0:00.00 on essentially every level. It's more a question of just how much "skill" it removes from the result, not so much a % of RTA vs IGT. In this game, at least, you still have to execute well to get a good time, even if the resulting distinction looks tiny in some cases.

In general, though, yes, that's the whole reason we make the separation to begin with.

[quote="zastbat"][quote]They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.[/quote]

iirc watching certain cutscenes is better since for whatever reason enemies killed in them are counted towards lowering the IGT which would fall under manipulating IGT.  although I have not watched all of the ILs so I have no idea if cutscenes that did not do this were left in.[/quote]

Right, I'm not referring to that, and neither was the verifier. This was a different cutscene.

ridd3r.

2014-05-19 12:28:36 pm

we have lift off

Quote from zastbat:

Quote:

They gave me a couple examples of levels that were unacceptable (watching the cutscene was one example) but unfortunately I wasn't able to get much detail out of them, mostly general comments about how the overall quality needs to be higher.

iirc watching certain cutscenes is better since for whatever reason enemies killed in them are counted towards lowering the IGT which would fall under manipulating IGT. although I have not watched all of the ILs so I have no idea if cutscenes that did not do this were left in. Someone can correct me if im wrong I might be remembering incorrectly.

Comic strips were deliberately played out, people killed who didn't need to be and had no effect on the timer (thus resulting in a slightly indirect movement route) etc. It was blatant taking the piss, which is basically why this submission got rejected. The runner himself even stated in a post that if the run gets accepted then it's a joke!

AdamAK

2014-05-19 02:54:48 pm

Totally rad

Quote from UraniumAnchor:

I don't really have a hard rule for this, but as an example of something we decided not to accept, one of the Sonic Heroes teams (Dark, I think?) has a move that freezes the game timer and lets you get 0:00.00 on essentially every level.

Just out of curiosity, why is a run like Pokemon Yellow accepted then? You get 0:00 IGT regardless of how well you play anyway.

UraniumAnchor

Edit history:

UraniumAnchor: 2014-05-19 03:09:49 pm

2014-05-19 03:08:27 pm

Not a walrus

Quote from AdamAK:

Just out of curiosity, why is a run like Pokemon Yellow accepted then? You get 0:00 IGT regardless of how well you play anyway.

You could make the argument that if somebody wanted to "improve" the yellow time by getting another 0:00 that was faster in real time, that could constitute obsoletion. In practice I don't know if anybody wants to bother.

The key difference in Heroes, if I remember correctly, is that you could otherwise play the game normally while the timer was frozen, so the run was effectively no different than not using the move other than an annoying "cutscene" every 20 seconds.

PJ

2014-05-19 03:14:32 pm

Is PJ

Quote from UraniumAnchor:

Quote from PJ:

I agree with Kotti's post. If the category is IGT, then manipulating IGT should be the main goal.

I'm not sure when this was ever in debate, so I apologize if anything I said misled people. Is there something I should clarify?

That was based on my interpretation of this:

Quote:

1) It is fine that the spin reload freezes the timer, but that does NOT mean you can just spin reload endlessly without a good reason. I'm not sure this would do you any good anyway.

If it freezes the IGT (which would result in a lower IGT), then I'd think that they should do it whenever possible. Your first rule seems to prohibit that, though. Same with the 3rd rule, which makes it sound like detouring to kill extra enemies to lower the timer would be against the rules/category/whatever. I was just clarifying that I'd approve of whatever means are necessary to complete the level with a low IGT, even if it is absolutely horrible to watch.

UraniumAnchor

2014-05-19 03:18:39 pm

Not a walrus

Quote from PJ:

The IGT (which would result in a lower IGT), then I'd think that they should do it whenever possible. Your first rule seems to prohibit that, though. Same with the 3rd rule, which makes it sound like detouring to kill extra enemies to lower the timer would be against the rules/category/whatever. I was just clarifying that I'd approve of whatever means are necessary to complete the level with a low IGT, even if it is absolutely horrible to watch.

Regarding the first rule, here's a non-exhaustive list:

"Good reason": There is a timed event happening and you want to freeze the timer during it.
"Bad reason": It doesn't cost any IGT so I'm just going to spin here for five minutes because I can.

The third rule is more that I might have to make an exception if some level has infinitely respawning enemies, because then you could just sit there for as long as you like, keeping the timer at zero. I don't think this occurs anywhere in the game, but I was trying to think of an extreme example.

[quote="PJ"]The IGT (which would result in a lower IGT), then I'd think that they should do it whenever possible.  Your first rule seems to prohibit that, though.  Same with the 3rd rule, which makes it sound like detouring to kill extra enemies to lower the timer would be against the rules/category/whatever.  I was just clarifying that I'd approve of whatever means are necessary to complete the level with a low IGT, even if it is absolutely horrible to watch.[/quote]

Regarding the first rule, here's a non-exhaustive list:

"Good reason": There is a timed event happening and you want to freeze the timer during it.
"Bad reason": It doesn't cost any IGT so I'm just going to spin here for five minutes because I can.

The third rule is more that I might have to make an exception if some level has infinitely respawning enemies, because then you could just sit there for as long as you like, keeping the timer at zero. I don't think this occurs anywhere in the game, but I was trying to think of an extreme example.

PJ

Edit history:

PJ: 2014-05-19 04:18:41 pm

2014-05-19 04:18:30 pm

Is PJ

Having not actually seen the ILs, that makes a lot more sense. Thanks!