Omok Score

Can anyone tell me how it's calculated?

 Badly

In all seriousness, I can tell you that the opponent can't be more than 500 underneath your score if you want to lose points.

Also, if anyone actually loses to me, they lose about 200 points.

I hope your losses are intentional... I've never even seen a person with a negative score. :[

I hope your losses are intentional... I've never even seen a person with a negative score. :[

-1337...looks intentional to me!

Uhm, I'm not sure of the difference but I believe it goes by 100 brackets. For example, if its 2100 vs 2200 you'd win 12 points but when you lose it's also 12. Doubling everytime you lose/gain. Now if theres 100 point difference it halfs everytime it seems. For example, I have 2500, if I play someone with 2200 I'll only gain 3-4 points, depending on the 10th's. However the bigger the gap also means the bigger the loss if you lose. If I lost to someone who was a clean 0/0, I'd lose 36 points, vs gaining 1.

These arent exact or even completely correct answers but they give you an idea.

I was looking for a direct extractation, but if it'd help to have data, I should collect some. I'd appreciate if Southperrians can help.

Seriously? Nobody to help me out?

Game	Player 1		Player 2		Who Started?	Score Difference
1	2463	0	2498	0	1	35
2	2474	+11	2487	-11	2	13
3	2465	-9	2496	+9	1	31
4	2476	+11	2485	-11	2	9
5	2496	+20	2465	-20	2	31
6	2485	-11	2476	+11	1	9
7	2494	+9	2467	-9	2	29
8	2484	-10	2477	+10	1	7
9	2493	+9	2468	-9	2	25
10	2483	-10	2478	+10	1	5
11	2492	+9	2469	-9	2	23
12	2482	-10	2479	+10	1	3
13	2491	+9	2470	-9	2	21
14	2481	-10	2480	+10	1	1
0	2490	+9	He DC'd	N/A	0	N/A

Someone please find out how to make sense of this data. The above table consists of consecutive games. I started the first game.

As you can see from this table, you gain a lot more points winning when you don't start vs. when you start. Also as you can see from the table, the points pretty much equaled/canceled out in the end.

In all seriousness, I can tell you that the opponent can't be more than 500 underneath your score if you want to lose points.

Game	Player 1		Player 2		Who Started?	Score Difference
1	2490	0	2000	0	1	490
2	2490	+0	1999	-1	2	491
3	2492	+2	1995	-4	2	487
4	2494	+2	1991	-4	2	503
5	2496	+2	1987	-4	2	509
0	2498	+2	1983	-4	0	515

The cutoff is definitely not 500.

Game	Player 1		Player 2		Who Started?	Score Difference
1	2502	0	2000	0	1	502
2	2502	+0	1999	-1	2	503
3	2504	+2	1995	-4	2	509
4	2506	+2	1991	-4	2	515
5	2508	+2	1987	-4	2	521
0	2510	+2	1984	-3	0	526

Notice in this second set, I also played 5 games, but in the end, the second player only lost 3 points and not 4 like in the set above, even though I gained the same 2 points.

The point gain/loss seems to be off by one row. The +/-s should be moved one row up for the chart to read correctly (which would explain the 0 under Who Started? on the bottom row of each chart).

From the data that's available, you get 10 points if you have less than 30 point difference and 11 if you have above. You also gain 9 if you have anywhere from 1 to 13 point difference.
You gain twice as many points if you win a game hosted by someone else and lose twice as many points if you lose a game you hosted.

That's all I could figure out from the data. If you want to figure out the cut-off lines you should have two fresh characters keep winning/losing and keep track of the points gained/lost and the score difference. If you want to see if the absolute point value has anything to do with the formula get a third person and lose against them on the character that has been winning in the first experiment. Once you shrink the score difference back down to where you have the data, run the experiment again and see if the results are the same. The whole experiment should take less than an hour if you can control 2 computers (which I can't b/c otherwise I would've collected data for you before). Still nice work with the ground work

I was going to post something like this in the middle of last week, but then my browser crashed and my entire post went with it. Anyway, here goes:
- - - - -
Your (thus far) limited data kinda fits an ELO model -- commonly used to calculate player ratings in two-player games. So does your data fit such a model? (The rest of this post explains how it might fit.)
- - - - -
Consulting this for reference: http://en.wikipedia.org/wiki/Elo_rating_...al_details
The rating change for a win or loss is RA' - RA = K (SA - EA). Furthermore, SA = 1 for a win, 0 for a loss; EA = probability (percent chance in decimal) of winning and the formula for this is given. So the big unknown would be that multiplicative factor K.

1) Your data from Table 1 seems to indicate: K of 20~21 (for both players) if the winner made the first move. Possible K of 40~42 (again for both players) if the winner was second to move. (This case happened only once in this data set.)
2) Data from Tables 2 and 3 seems to indicate:[INDENT]Player 1: Moving first - K of 20~21 is still a decent guess/estimate. Moving second - K of 40~42 is a decent guess/estimate.
Player 2: K values seem to be roughly double of what applies to Player 1 (IIRC, roughly 40~42 and 82~84 -- remember: I originally wanted to post something last week). As mentioned in the Wikipedia article, it's common to apply a different K to players of different ratings or when total number of games played (i.e., number of samples) is low.
[/INDENT]So going forward, you want to obtain data that:[INDENT]a) proves or refutes K of 20~21 for Player 1. And/or find the precise value of K.
b) shows if a 2x multiplier is applied when the winner was second to move.
c) shows what the different rating brackets are.[INDENT]Before the possible 2x multiplier:
It looks K = 20~21 for 2463~2508. --> rating bracket is 2400+? Or is the floor for this bracket something else? Is there a ceiling to this rating bracket?
K = 40~42 for 2000 and below? Or what is the floor/ceiling for this bracket?
[/INDENT]d) Are there different rating brackets for different total number of games played?
[/INDENT]NOTE: IIRC, assuming the game rounds rating change down, up, or to the closest integer didn't exactly match when I specifically tried K = 20, 20.5, or 21 to your Table 1 data. (Thus, I specified a possible range for K.) I'd have to double-check this though; or you/someone else can. Additionally, more data and targeted data could help pinpoint the precise K value(s).

According to the Wikipedia article, the calculation of the probability of winning, which is also factored into the whole equation, changes depending on previous games. Is there a base formula that works so that the K value is the only missing multiplier?

More results:

Game	Player 1		Player 2		Who Started?	Score Difference
1	2501	+6	2394	-6	1	107
2	2507	+15	2388	-15	2	119
3	2522	-16	2373	+16	2	149
4	2506	+5	2389	-5	1	117
5	2511	-15	2384	+15	2	127
6	2496	+6	2399	-6	1	97
7	2502	-14	2393	+14	2	109
8	2488	+7	2407	-7	1	81
9	2495	+16	2400	-16	2	95
10	2511	+14	2384	-14	2	127
11	2525	-16	2370	+16	2	155
12	2509	+5	2386	-5	1	123
13	2514	+14	2381	-14	2	133
14	2528	-16	2367	+16	2	161
15	2512	-24	2383	+24	1	129
16	2488	+7	2407	-7	1	81
17	2495	-13	2400	+13	2	95
18	2482	+7	2413	-7	1	69
19	2489	-13	2406	+13	2	83
20	2476	0	2419	0	1	57

At no point did either my opponent or I leave the game until the end of this table.

It might be easier to find that debug file and extract it yourself.

Not there, and I didn't see it in the BMS server data either. Gotta be one or the other. Good luck searching for it because I had a hell of a time doing it.

Not there, and I didn't see it in the BMS server data either. Gotta be one or the other. Good luck searching for it because I had a hell of a time doing it.

How the hell did you get the serv...nvm, so the formula's neither client-sided nor server-sided? How is that possible? Hard-wired into the executable itself?

How the hell did you get the serv...nvm, so the formula's neither client-sided nor server-sided? How is that possible? Hard-wired into the executable itself?

It's not missing, just very hard to find in an unlabeled exe. You have to look through several MBs of code in either case. The easier way to get at it would be to debug while MS is running and see what happens when you win a game. But you tend to get banned for doing that.

Then how was the magic formula extracted? How about monster damage (PDD)?

Then how was the magic formula extracted? How about monster damage (PDD)?

I remember them being labeled like "mdamage_PDD" or something.

Maybe you could run a private pomegranate and debug it while it's running?

Yeah that's your best bet, or ask an owner to do it for you.