blocking odds

[Glamdring]

Thanks to Snarr for the sticky :)

It was brought up in the RD thread in GD that once you had successfully hit a knight from the side, you could continue to do so without fear of blockage, as long as your unit didn't move. In response to this preposterous suggestion from :SLAP:NASTY:, I did a little grey-on-grey Testing with two new accounts and discovered to my surprise that he was completely correct

The results of the Test were as follows:

once an attacking unit hits a side-facing unit, he does in fact continue to successfully hit with the same attack
any attack by any other unit on the same piece will interupt the hitting streak
regardless of whether the attacked unit blocks it or not
healing does not interrupt the streak
moving either piece will interrupt the streak
changing the direction the attacker faces will not affect the streak, but
changing the direction the attackee faces will
the results are equal and opposite when the attack is blocked
the results are equal and more spectacular in a frontal assault

[Snarr]

Like I said earlier this is really big, I can already imagine myself hitting a knight on the front 10 times. In fact that does explain how my knight got hit 5 times in a row on the front.

You didn't even failed to credit :SLAP:NASTY:, which is admirable.

I dub thee the biggest discovery since the advance scout LOS.

One must wonder now whether this is intentional or not.

[LondonJack]

Just tested this with Jeffery on Banff, and we had two failed attempts to start a streak of Knight frontals. (I was grey, Jeffery was gold)

I was going to test the effect of friendly focus on the streaks, such as the BW and Stonie, but couldn't seem to get a streak going.

Not sure if the Stone is to blame here, but in the first failed attempt, both the attacker and attackee were stoned, and between hits the attackee was blocked by the attacker. Second failed attempt, the attacker was stoned, but not the attackee, and the attackee was healed between attacks, but did not attack and was not attacked otherwise.

Edit: there does seem to be something to this though, on another test grey vrs. grey, i was able to do 4 straight frontals on a knight, but the 5th attack was blocked (the attackee was not manipulated at all during this time, but was healed several times)

[Glamdring]

Originally posted by LondonJack
both the attacker and attackee were stoned

maybe that was the problem... drugs are bad, mmmkay?

[LondonJack]

From a conversation that occured after the stoned Pink Knight blocked the stoned Blue Knight.

Pink Knight: Hey man, you were supposed to hit me remember?

Blue Knight: Aww man, i forgot.

*long pause as the turn clock winds down*

Pink Knight: Are you going to change your facing or what man?

Blue Knight: Woah man, i thought it was your turn man.

Pink Knight: Nah man, not until you face or use the hourglass man.

*Pink Cleric heals*

Blue Knight: Wait man, why did you block me then man, if you knew i was supposed to hit you man? Let's try it again man.

Pink Knight: OK lets give it another try, this time i will take my helmet off, and lay my shield down.

Blue Knight: Wait man, i can't do it this turn man remember, i have to recover, pass me that pipe while we wait man.

Pink Knight: Hey man don't bogart that man. Dude, we so need to recruit a plumber into our armies man.

Blue Knight: alright man, I'm ready now, take off all your armor so i can get a good shot at your head man.

Pink Knight: alright man, but don't laugh at my poka dot boxers man

Blue Knight: Your wearing Pink armor man, how much worse could your boxers be man?

[Tcbb]

Please post your test results and statisics here. Only numbers can explain things.

My 15 experiments(more needed if time allow):
f 0 3. (attack from side f ailed. I attack again and suceed (in attacking from front)0 times in a role. This happen in 3 of my experiment)
f 1 3
f 2 2
f 3 0

s 0 2
s 1 4
s 2 1
(I didn't heal, so by the time the knight receive 3 hits it dies)


My theory is that after a side attack fails, and we attack from the same place, which is a front attack now, it still counts as a side attck.

It's easy to confirm this bug. Xerent had the source code, I believe. I could download a flash dicompiler, but I can't install it.

[Agent Smith]

I'll need to test this, but I'm pretty sure that I've missed a front shot on a witchwith my scout, not had either peice move, and then hit from the same spot next turn.

I'll see if i can look into that.

[xerent]

Ugh, I do NOT want to look through 6,931 lines of code. ^^;;

[Tcbb]

Could you send it to me via e-mail? Based on the example you showed, i think flash is not much different from C++. Thanx.

[xerent]

For that, I would require the use of an e-mail address. ^^

EDIT: Also, I would like to get an ok from DR... technically, this is against Terms of Service... even though our intentions aren't malicious, but the opposite thereof.

[Tcbb]

Thanx, my e-mail address is tcbb7@yahoo.com.
I think you didn't pm DR when you decompile it:) wrink, wrink ;)

[FryLock]

This sounds incredible...but perhaps it only applies to melee damage? As Agent Smith said, I am certain that I've had a witch block my scout, only to hit her 2 turns later. In fact, it just happened. So maybe it's just with 1-space attacks?

Hmmm....

[Elentari]

hey... maybe it only works for greys :).

[Glamdring]

or maybe I was wrong about the blocking, and right about the hitting... I didn't test the blocking part quite as extensively...

I will, of course, look into it again. :cool:

[Tcbb]

This bug definitely doesn't exist in Legends:
After a side block, it takes my 7,10, 17,18,21 hits to kill five knights.

Seeds gave us an inferior version!!!!

Man Glamdring, you have good observation. But you definitely need more research before you publish your paper :) Even my examples are very insurficiant...

[Aro23r]

I tested this out myself. I kept attacking the same knight from the frong with different pieces until one of them hit. (Turned out a knight did the damage, eventually.) I passed some turns, until the attacking knight was able to move again. I attack again, without moving or anything, of course, and the attackee blocked.

According to your blocking 'rules' this should not happen though, correct?

[Tcbb]

You should attack with one piece.

You need more data to support your case.

I am currently doing more experiments, and I get some confusing results. I will try to come to more conclusion.

[Tcbb]

My new data seem to contridict my previous obsrevation.

On Armegeddon:
After a knight side block.
It takes 14, 17,21,21 hits to kill 4 knights.

On Banff:
takes 7, 9, 19.18 hits to kill 4 knights.

my previous result has about 0.5% chance of happening if THERE IS A BUG.

My resent result has about 0.1% chance of happening if THERE IS NO BUG.

All I can say right now is that there is a big derivative.

[abyaly]

I tested this with my home-made wooden TAO board and a set of dice. so far glam's theory is proving correct.

so we know that:
on Army, glam is right
on Banff, glam is wrong
on abyaly's home-made TAO board, glam is right

who tested on GL?

[Mnrogar]

I did a little testing of my own on army. Grey vs Grey. I had a knight run around hitting another knight of my other account, each time knight 1 missed I'd move both units around and then do it again. Once I got knight1 to hit knight2 with a front hit, I didn't move either and did nothing but skip turns and heal until he could go again, when he did from the same spot still trying his front hit it missed. Knight1 was the only unit that attacked Knight2.

Then I tried side hits with the same result; when I got a hit the second try it missed.

Bottom line, I got the same results as Aro.

[Tcbb]

Please do at least half an hour of testing before you post,
I read pretty slow....

Anyway in statistic point of view, it doesn't matter if you are using wooden die or what; doesn't matter if your knight ask for mercy before it dies.

Let's get serious on this issue.

[abyaly]

the dice arent wooden. they're plastic. a translucent shiny red plastic.

[Mnrogar]

Please do at least half an hour of testing before you post,
I don't know if that was directed at me but I don't see the point of going on testing for 25 minutes if you can contradict the theory in the first 5 minutes.

[Glamdring]

Well... my test was done on army... I wish there were more consistent results... this is confusing :confused:.

[plusminus]

I have seen this happen over a few games, on GL, with scouts. Very consistently. Perhaps it only affects certain units?

Just happened in the last game I played, and the scout fired an arrow (but didn't move/change directions) in between hits. The attacker, a knight, got a front hit, retreated down the board, and then went back up for another front hit.

[FryLock]

Although, it is possible that what you describe could happen with a scout...60% blocking is good, but it's not outside the realm of (understandable) chance that a scout could get tagged from the front twice in a row.

[Office_Shredder]

yeah... 60% blocking means that almost 1/5 of the time you will get hit twice in a row from the front.

[Only_Option]

You guys are like little nerdy scientist ;p.

[FryLock]

*puts on his lab coat, tosses Office a pair of protective goggles, and pours sulfuric acid on only_option's head*

You were saying...?

[iron-rain]

This stuff didnt really work out for me.

[*The Nothing*]

Yeah i tested it as well iron-rain, it didn't seem to work for me at all.......

[QuicksandSlowly]

This is very interesting stuff. I'm really not willing to spend any time experimenting with this all, but it seems to me that most of you are just testing this stuff on front hits. Maybe Glamdring was talking about side hits... :confused:

[ashley/hulky]

what happened to the sticky?

[Glamdring]

It was unstickied because the inconsistencies were never solved. If the theory is bunk, I'd actually rather it weren't a sticky anyways. But I really should do some more testing, now that I think about it... I've got a little free time at the moment, maybe I'll work on it.

[Stones]

i have tried this and it did not work

[Darrington]

This interests me greatly and I wish to have some solid consistent evidence.

[Glamdring]

I did some more testing and I think I may have something. Most of the time, the theory was perfect. However, once every five minutes or so, something would happen that directly contradicts it.

So this is my new theory, one that is unfortunately impossible to prove without deep and intimate knowledge of the game programming. I think it has to do with the randomizing; somehow the value (hit or miss) doesn't change unless some other move is made. However, the random code that is generated is regenerated every 5 minutes or so, so that it doesn't stay the same.

If that starts looking like ancient egyptian hieroglyphs the longer you stare at it, I'll summarize. It works. Most of the time. Then sometimes it doesn't. So don't bet your life on it, or more precisely, don't bet your last scout's life on it.