Roll Eyes
Lips Sealed
<- 1234 ->
List results:
Search options:
Use \ before commas in usernames
Edit history:
LotBlind: 2012-04-11 08:18:34 am
Great job noticing that Rane! Yeah difficulty seems to stay low if you keep getting rank E's. I wonder if that means getting extra rank E's is worthwhile though? Or is better to ignore the other zombies during missions? I'd like to think it's better to get rank E's even if they're extra, because I'm guessing the game tracks the average rank and not the body count.

In my experience and after a bit of testing just now I'd say an accurate/inaccurate word typed in the same time gives you the same rank. What someone should do is get their friend to mash buttons randomly while you're typing as usual, and then repeat the same chapter without that. The effect, if it exists, should be noticeable by then. I also don't think it makes a difference what your accuracy was on the performance report sheet vs. how hard the next chapter will be. This can be tested in a similar way.

Has anyone thought about if it matters in which order you type the zombies in a situation where you have the time to do it in any order? If you follow the camera panning left and right according to where the zombies are, you'll see the last zombie left will be centered on when the second to last is dead. After he's down, the camera eventually returns to a sort of middle position, the position it was in when those particular zombies first appeared. Question is: is the time it takes for the camera to return to that middle position (after which the game seemingly starts to follow pre-scripted camera movement instructions) dependent on how far the last zombie was from that position? I hope I've expressed this clearly enough.

If it is, and if you wanna be a perfectionist, you might it into consideration at some points, and kill off the one closest to that position last. But in any case this could use some testing.  The results from my initial tests aren't as conclusive as I was hoping so I'm not sure.

This sort of stuff is why, if you ask me, an IL run would be smarter to start off with cause it just reduces the amount of variables by a lot while we're still looking into these things.
I'm just trying to impress you.
Is there any way to choose the alternate chapters in IL's? Because alternate 3 and 4 are much faster than their default counterparts.
Really don't think there is, but that's not really a problem. I guess if there was you'd have to choose the alt ones if they're faster.
Just a brief notice that I'm doing testing on the camera pan/which zombie to kill last -thing, and I already have some findings, however there's too many uncertainties still so I'm going to put off presenting them for another week or so.

Cheers everyone!
Wouldn't it be more interesting with 100% accuracy IL runs? I'm interested in running this but the whole type-wrong-to-make-the-game-easier thing is kinda off-putting.
I want off the ride....
Quote from Freezard:
Wouldn't it be more interesting with 100% accuracy IL runs? I'm interested in running this but the whole type-wrong-to-make-the-game-easier thing is kinda off-putting.

you can still get rank e's with perfect accuracy.

the point of missing is to reduce your HP totals and to make it easier to plan deaths. if you try to account for 100% perfect typing and those HP ins and outs. your stuck planning for a run that will NEVER happen. There are too many random phrases that come out and a lot of them have to be typed too quickly.

A rank "E" is just meaning that you didn't type it quickly enough/within a reasonable level of accuracy. Best way to prove this is to setup an auto Q macro and type everything out. You'll probably still get some A's and B's but you'll generally (on the longer phrases) go down in rank severely. Now would you wanna do this in the run? Not so much. The point is to abuse the fact we take damage the general fastest way (taking an axe to the face) and in this time we are generally stuck waiting anyway. Making the game easier by ranking "e" on the axe thrower... is well just a convenience. SOooooooo

SOOOooooooooooo if your really upset about "type-wrong-to-make-the-game-easier" thing. Then you should try a run of 100% accuracy. I am pretty sure you wont have a solid time getting out of 3 stages. And an abysmal one finishing the run. Outside of magician where 100% accuracy is well a requirement. More times it just is messy to be 100% accurate...

the fact that HP loss can save SOO MUCH TIME, and that time is mostly saved in skipping cutscenes by game overing, or taking a game over in ab oss to a. get a free hit, b. reduce the difficulty so its not OMFG hard.

Fun fact, difficulty just determines how fast zombies hit you. So if you TURN UP the difficulty, you just need to type faster.... which in turn gets more rank a/b's and makes the game harder for you.

Score runs are also painful for the same reason.
Erm I'm not upset about it, just too lazy to plan out all sorts of possibilities when taking damage, but it looks like you already have it figured out Smiley I saw PEACHES run and it's looking good!
If anyone wants to help with figuring out how this game works/helping with the run that someone's bound to post eventually (which I think has already benefited from all this investigation me and others have been doing), PM me and I'll explain what you can do. Fast typing is a boon, but not necessary. And also is anyone interested in sharing thoughts with me about what research the running would most benefit from?

BTW guys, what's the deal with PC games and variable FPS? Whenever I'm running this game, the FPS might vary from moment to moment right? Oh and is there an ongoing debate about the possibility of slowing a game down in order to cheat? I mean if someone was running this game with a older computer, even just the extra CPU usage from the software that's doing the capturing might slow the game down mightn't it? And do PC runners have to submit a "proof tape" showing them actually, say, typing away their run of TotD?

What rank you get only depends on how quickly you type a word starting from when you've typed the first letter. I'm pretty sure of this now that I saw that's what the manual has well.

IMO 100% accuracy would make some sort of "superplay" video, but not a speed run.
At this stage I seem to have overcome a big problem (by gritting my teeth and... the problem turning out not to exist) and am able to thoroughly analyse the flow of the game like I wanted. What I would really appreciate now is if I could take a look at the routes the runners (peaches I guess) are planning to use, i.e. which civilians are saved etc. so I can save a lot of time sticking to the routes that have been found fastest. Of course if there's some doubt, I could look into the alternatives being weighed.

Also just downloaded Hourglass and gonna see if it's going to be useful here.

PS I'm pretty sure now it doesn't make a significant difference which zombie you type last in any attack, because the camera pans faster if it has to move through a larger angle.
The FPS might vary, but a semi-modern PC should be able to capture this game at a stable 60 FPS. Lower FPS doesn't mean slower gameplay, the time still continues in the same speed.
Edit history:
PEACHES_: 2012-04-28 03:03:20 pm
I'm just trying to impress you.
I have a really powerful PC that still drops frames at times while capturing (most noticeable is the glass breaking at the start of ch1)

I was holding off on doing more runs until a determination was made about zombie order. Now that it's not an issue I'll try to start recording again soon. Here's my new route if you want to test things within it:

Ch1: Save the civilians at the fountain, let the woman die to the barrel
Ch2: Save car civilian, skip the key, save woman before gate, let all others die, continue on the fish
Ch3: Open Water Gate, let final civilian be attacked
Ch4: Continue on mission, save civilian in vent
Ch5: Continue before car ride, continue on Magician
Ch6: Continue on second mission
Peaches, what do you meam when you say "continue on"?
I'm just trying to impress you.
"Continue on" means lose all your lives and spend a Continue. I find those are the spots where the massive difficulty drop from using a continue are most advantageous.
Quote from PEACHES_:
Has this been abandoned? I picked up the game and intend to collect the bounty. Here's my skeleton route of major choices:

Ch1: Save the civilians at the fountain, continue on the mission, let the woman die to the barrel
Ch2: Save car civilian, skip the key, save woman before gate, let all others die, continue on the fish
Ch3: Open Water Gate, let final civilian be attacked
Ch4: Continue on mission, save civilian in vent
Ch5: Continue before car ride
Ch6: Continue on second mission

I still need to optimize intentional errors to avoid extra HP from filling up the perfect gauge, from there it's just RNG grinding a good run in terms of agreeable phrases and tranquilizers. With that route I should be able to cut 30 minutes any% ss on 3/5/Normal. My final stats tend to hover around 30 in-game WPM.

I actually haven't abandoned this. Just unfortunately haven't had the time to work on it like I would like. My first run of this was completed about a month ago and got lost in a hard drive crash. Got pretty discouraged after that for awhile. Sorry about that guys.

I'm glad you're picking it up too. Your route is actually looking strikingly similar to mine which helps reaffirm what I was thinking. I'm looking forward to seeing what you pull off with this.
I'm just trying to impress you.
Do you recall what your completion time was?
So are you saying you're no longer gonna pop a continue in ch1 at all? Oh I see, you've saved that one for magician. Makes sense. How about crates? You probably already figured out which crates you wanna smash? Like in Ch 1 the first crate will sometimes give you a tranq, sometimes balloon head meat, so no reason to skip it. The second gives you either tranq or zombie suit so again you should type it, but in other places typing a crate can have mixed effects, positive or negative. If you have an even more detailed route, I'm always interested.

BTW in IL runs I think timing should start from when you hit enter to select a chapter and end when the last word has been typed, i.e. not counting the end cutscenes or performance sheet. But in SS I suppose you have to time those as well, so you can't afford to get a high score of any sort (top 5 or top 10 or what is it?) if you want to avoid the "name entry" bit. With all the rank E's etc. I guess you won't. But that means your time depends on out-of-game variables, the previous results from everyone who has played your copy of the game, so I don't know if that's wrong in principle. So maybe we just don't time that stuff. It's just enter-mashing anyway. Not very interesting. But then it is an SS... Oh yeah and a top 5 or whatever score gives you an extra life as well. We have to figure that out.

Good to hear from you Tetsu. Bad luck with the crash! Well we're in no hurry. I find it easy to think up the theory of how to study the game and very slow to actually get results, mainly because I've never been this scientific in my life Wink And it's just a lot of work. But it's fun too. And once again no need to get it perfect the first time around. A first submission can always be improved later on.
Edit history:
PEACHES_: 2012-05-06 04:31:27 pm
I'm just trying to impress you.
I'm not really interested in IL runs. Boss rush, maybe. But that's way more RNG dependent. Phrases tend to average out over a whole run.

My runs are hovering in the mid to low 32's, timed from hitting enter on "Arcade Mode" on the main menu to finishing the final Emperor question. My goal is to cut 32 and I'll submit once I'm under it, unless T3tsuya has evidence of a faster time.
I'm just trying to impress you.
just got 32:01 on a messy run. it's gonna happen happy
Edit history:
LotBlind: 2012-07-04 03:41:30 pm
So congratulations Peaches for your SGDQ run of Typing! I missed most of the marathon and didn't even realize the game was in the schedule until a couple of weeks ago. I was just watching the run now. Epic problems getting it to run Smiley Was very impressed with the Magician fight in particular!

I actually have a couple of things that I've been wanting to post for some time now, but I've been putting it off until enough loose ends have been tied, meaning until I can speak with enough certainty about my new ideas and speculations. However I'd like to know what anyone is doing about the game right now, which might affect what I post and how soon I post it. I have a few comments on the marathon run as well which I will certainly post here alongside everything else.

For now can you tell me about one detail in the route you've chosen Peaches - in chapter 1 did you or did you not mean to let the man die who's being assaulted by a zombie from either side? I'm talking about the one after the first crate, with his back against a well.
Edit history:
LotBlind: 2013-05-25 01:39:39 pm
LotBlind: 2013-05-25 01:28:41 pm
LotBlind: 2013-04-04 05:27:24 am
LotBlind: 2013-04-04 05:25:40 am
I wonder if you have a more recent run Peaches? I'm gonna dump some old notes now that were mostly made around the time I posted the last message on this thread. Some of this is more relevant than other things, some things you'll already have figured out, some things are kind of speculative, but maybe something will come out of it nevertheless. Sorry the structure is a bit sketchy.


Difficulty levels: I've almost certainly established the following. Every zombie in every situation will have a difficulty level, officially called rank, of its own. We know a zombie's rank in every case has a pre-set minimum, which differs from chapter to chapter and situation to situation. Obviously a zombie that you are given a long time to type will have a higher minimal rank than a zombie you must type fast. Fast typers, especially those who attempt deathless runs, will no doubt become familiar with maximal ranks, which I have so far observed with at least Magician (in Peaches' SGDQ 2012 run).

Zombie ranks fall strictly into certain quantum categories. This is, presumably, because all the words and phrases the game has in its dictionary are pre-divided by their rank, which roughly corresponds with their lengths. Rank is affected by the typing ranks the player is getting (A and B probably increase the rank, D and E probably decrease it, C possibly doesn't do anything), and by other difficulty-altering things.

Here are all the ranks up to 9:

rank                avg length         range of lengths*        example***
1                    1                      1                                  bats, axes and crates
2                    2 (2.5)              2 (2-3)**                     the first 5 zombies in chapter 1
3                    4                      3-5 (2-5)                      the next few zombies in chapter 1
4                    6.2                    4-8
5                    8.3                    7-10
6                    10.5                  9-12
7                    12.9                  12-14
8                    15.4                  14-17
9                    17.5                  16-19

*Note that special characters and numbers are counted as two characters each for the purpose of establishing the phrase rank. The numbers in parenthesis take into account singular or very rare cases where the mostly predictable pattern was broken.
**Always either two regular characters or one regular and 1 special character.
***Based on studies with the normal difficulty setting.

I believe the ranks go up to something like 34 characters for normal enemies (at least the guy in the elevator, chapter 6), but Strength can have longer sentences than that. There might be another exception somewhere.

As you can see, the categories can overlap. Therefore it is sometimes impossible to determine the exact rank of a single zombie without comparing with other samples acquired in similar circumstances. Luckily zombie attacks tend to feature multiple zombies of the same rank, or the same zombie can attack you more than once (e.g. Randies and bosses). In the case of some bosses rank can change dynamically, but in other cases the rank seems to be locked all through a boss fight or at least a phase within it.

The general tendency is clear enough: starting from rank 3, every successive rank has avg length increasing by a little over two. Every rank upwards of 2 has lengths that fit within a 3 or 4 character range. Mostly the range is 3, but exceptions can be seen. The ranks above 9 start to show more erratic behaviour due to partially unknown reasons. One obvious reason is my body of data slowly peters out the higher we go leaving a greater margin of error.

Maximal ranks:
If you watch Peaches' SGDQ '12 run and go to where he's fighting Magician, you'll notice that the phrase lengths don't increase at all during the fight. That's the maximum for Magician. The word lengths are 18-21, which makes it rank 10 or 11. There might be a bug in the game that makes rank 10 the same as rank 9 or then there's a bug or a quirk in the "Mr. B" drill (more on that later).

Because of the fact that there are minimal and maximal ranks, the decision to use a continue should be carefully considered lest the full use is not availed of. They must also be taken into account when testing for the effects of different methods of altering difficulty, so as not to e.g. attempt to show the effects of using a continue when the next few attacks are already at minimum difficulty.

So where am I getting this from? There is a drill in the game called "Mr. B Man". I was looking for a way to start studying difficulty systematically, when I noticed that this particular drill seemed to behave in a way almost ideally suited for my purposes. It gives you words of gradually increasing length to type with the aim of typing them in a rank A or rank B time. Failing to type fast enough ends the drill. Making a record of the phrases presented to me within the drill, I soon found... everything you saw in the table above. Obviously, the further the drill would progress, the more likely I was to have finally made a mistake, which explains why my body of data on higher ranks is limited.

Of those elusive higher ranks I can only tell you this: many more rules seem to break. At most the range of lengths increased to 7 and averages in one case would not increase at all between ranks 9 and 10. We have to keep in mind though that all of this is based on assumptions, for instance, about how the "Mr. B Man" drill itself works. Maybe the programmers wanted the difficulty to scale non-linearly if the typer made it past a certain number of phrases. This is, however, the best model of difficulty categories that I could build based on my records.


-every zombie has a certain "rank" (as per the manual) that correlates with either how long the player has to type him, or how long his word is. Sometimes both.
-every zombie (or even zombie type) has defined minimum and maximum ranks
-some zombies always have the same rank
-rank categories can overlap
-special characters and numbers count as two characters when establishing phrase length


First three zombies of chapter 1 - You may have noticed that when they're down, the camera follows the last body melting into the ground. The thing is, the length of the wait between the zombie's death and the game proceeding depends on which death animation he goes through. There is a set of the same 5 animations for most of the taller zombies.

So the best animation you can get is one where the David (which is the actual name for that type of zombie) falls directly backwards with arms thrown wide to the sides, but all the other deaths have pretty much the same length except for one. The one where he first falls to his knees, then collapses forwards. It's about one full second slower than the next slowest, the knees-fall-backwards, and 1½ seconds slower than the first one I mentioned. So something to think about in IL's I suppose, cause you can still reset at that point should you get the unlucky knees-fall-forwards if you feel you'll need that extra second to push your time lower.

Now I'm not 100% positive on this, but I think the best way to lose a continue in chapter 1, should that be desired, is taking damage from the pair of axe-guys right before the mission. You know if you want to get a zombie to attack you sooner, you'll sometimes slow him down by starting to type him before he has started/finished his attack. You could also aim for getting rank E's from the pair. So I'm guessing you might want to type the 1st letter of the one on the right AFAP, wait till the axe is thrown and finish him AFAP, because that triggers the other zombie's axe-throw. This time you'd probably wait for the axe to be thrown first, then start typing him slow enough for the E, but before the second axe lands.

Now here is where I was actually a bit surprised with your marathon route, but maybe you wanted to save the guy at the well (in the area after talking to G), but got distracted. The route where you save him is obviously faster.

So type those two zombies (at the well) quick, but the one after that is maybe a C or D or something if you're quick with the first letter, because of waiting for him to boot the civilian (takes around exactly one second).

Also you can get a low grade from the guy near the end of the level that's jumping down from a bridge, but you seem to know that.

time it takes for first mission zombie to bite you thus ending the chapter 1 mission: around 5-7 secs (from when the word first appears) depending on difficulty, although some missions (at least the ones in chapter 6) seem to have a fixed level of difficulty and fixed times as well. I just realized that you could type the other zombie that comes out for a rank E while you wait, or even a couple of them.
My intuition tells me it's better to get an optional low-rank kill (E or so) than not to get the kill at all, because I have a feeling the increasing or decreasing of difficulty depends on some sort of average of the ranks you've been getting as you play. I also suspect that that average is only calculated on an area-per-area basis, because I did some tests a while ago in chapter 6 where I would start the chapter with a given difficulty setting and proceed up to the elevator. All the zombies before the elevator had fixed ranks regardless of how fast I had typed the previous ones. The elevator one was the first to vary. I also have a feeling the game dumps all previous data on your performance, i.e. it doesn't matter how fast you were typing in the previous area in terms of what happens to the difficulty level after the current one.

time saved by using up a continue in a mission vs. getting a regular mission fail: 3 secs

Not only do you lower difficulty when continue-ing, but it also hits the attacker, either killing them or damaging a boss. HOWEVER, a boss will only take damage if the continue was used on a damaging attack, i.e. not on Magician's or Emperor's "catch ma balls" attack.
Edit history:
LotBlind: 2013-05-29 04:44:22 pm
LotBlind: 2013-05-25 02:15:18 pm
Here's something maybe marginally useful.

I don't think it's been pointed out explicitly, so the rank you get for a phrase is determined by a timer that starts when you've typed the first letter and stops with the last letter. Reaction time (i.e. first letter) is not counted. This means that to get low ranks, it's good to start typing ASAP. The time is then, I believe, divided by the number of letters for an average time-per-letter figure which determines the rank. However you need to pause for one beat somewhere within the word to correct the bias created by the first letter - it was not timed, but it is included in the calculation. You also need to count numbers and special characters as two characters, i.e. either pause for one beat before typing them or skip that beat somewhere else within the word.

In case you're not sure why it's like this:
Imagine you're typing 2-letter zombies at a certain bpm x. Let's turn that around to minutes-per-beat y=1/x. It takes you 1 beat to type the last letter (first letter is not timed). The time it took was y, but that's divided by 2, because there were two letters. So we get y/2 minutes-per-letter. If you next type a 10-letter zombie at the same mpb y, it takes you 9 beats to complete the word and the time is divided by 10. So we get 9y/10, which is a lot closer to y than y/2 was. You can easily see the effect by typing the first few zombies in the game. And we can also see that if we add 1y to the typing time (add an extra "empty" beat), we get the expected y minutes-per-letter consistently, or y/60 seconds-per-letter.

I seem to have found a pretty consistent barely-rank-E tempo (if you use an actual metronome in your running) at around 210 bpm. At 250 bpm it seems to be pretty consistent D's. I wanted to do some more accurate analysis of test videos at one point, but I was recording with the wrong settings unfortunately. Maybe at some point.
Edit history:
LotBlind: 2013-05-29 04:35:28 pm
LotBlind: 2013-05-28 02:17:24 am
I have started a TotD Strategy Wiki. I will be adding more things from this thread and from my own "knowledge base" onto it in the future, but all contributions are appreciated. Editing pages is relatively simple, I had not done any Wiki editing before.

EDIT Here's the link:

EDIT And just to be clear, you don't need to know a lot about the game. If you like playing the game, you're probably gonna be able to contribute to the Wiki simply by recording your playthroughs, albeit in a coordinated way. Just contact me.
Edit history:
Shox: 2013-07-31 02:41:25 am
Shox: 2013-07-31 02:39:14 am
Sorry for bumping an old topic, but is there still a lot of interest in this game? I managed a 32:36 on my second run without a whole lot of optimization, and also with a lot of mistakes on levels 2 and 3. I still need to learn where certain items appear and when to intentionally lose lives. Thankfully there seems to be a lot of useful information here. Smiley

I really need to study Peach's run at SQDQ of last year a lot more just to see what he does. Does he still run this game by the way? I know he said he was going trying to break 32 minutes and he had a 32:01 run, but I never saw him post anything else. :/

It's a really fun game to play though!
I have the aforementioned project (strategy wiki) ongoing although on pause while another project is in the works. If you think you have anything to contribute to the wiki, go ahead, I could even tell you what's most relevant. And it's as easy as recording your playthroughs in a systematic way. So it's maybe not quite the same as speed running the game, but seeing as you obviously have typing skills... I'm not much faster than average myself.

I haven't heard from Peaches (or any other runners) in a long while, so unfortunately I'm not sure if they have information not posted here yet.

To give you an idea of the sort of thing that needs to be studied:
-what items can you get from what crates? missions? (what you already mentioned)
-how slowly you need to type for an E
-how critical is it to get those low word ranks where available?
-what's the best overall route(!)? (peaches and another runner whose run got rejected had different routes, albeit in slightly different categories, and it wasn't clear to me just based on their runs what route is fastest)

But a lot of these things can be studied more efficiently if you first list everything and then try to be systematic, so if you're at all interested, we could coordinate the effort.

And so there is an interest Smiley For me, a non-speedrunner, it's more to find out how the game works, but then that whole thing started by trying to help find speed running tricks so... they go hand in hand.
Edit history:
Shox: 2013-07-31 11:19:21 pm
Shox: 2013-07-31 08:03:23 pm
Shox: 2013-07-31 07:56:43 pm
I'll definitely update the Wiki in a little bit. Thank you for what you've posted in this topic, it's definitely helped me. It really is a pretty interesting game.

I feel like the route I'm doing is probably as good as it gets, I pretty much just copied Peaches route haha. I just got a 32:04.77 run with a ton of mistakes on the last boss, as well as quite a few slip ups on the third level boss (sorry I can't remember their specific names Sad ). I also didn't single cycle the boss in the second level at all, which is HUGE. It's here if you wanna see it:

As Peaches said earlier, sub 32 is definitely possible. I'm not too sure what time would be worth submitting as a run here, though?

Edit: Also, I'm going to openly admit that I was the runner that got rejected haha. I had no clue that so much research had been done on the game and I didn't have any specific route at all, the routes you guys have discovered here were much faster than anything I ever did. Smiley However, I'd like to submit another one with a lot of improvement for sure.