FAQ

pier4r
Skynet
Skynet
Posts: 2768

On score inflation for master league and how to compare historical scores.

Post#21 » 22 Jan 2017, 19:49

On score inflation for master league and how to compare historical scores.

With the current conditions, the higher score will slowly grow indefinitely if there is enough activity
it seems that there is an inflation of score points in the top league, maybe due to points taken from players joining the top league, losing points and then stopping playing or going back to lower leagues. In this way (at glance, it has to be formalized a bit better) the top league is slowly accumulating more and more points that then end up on the top players (especially if those do not get inactive for long time).

The phenomena is balanced only when top players (top league) stop playing, making points not accessible to others (because there is no exchange of points with those players).

Once reaching 1700 points was difficult, then reaching 1800 points was difficult, then will be 1900 and so on. Unless, as written above, top players go inactive and the points gained are removed from the pool of points exchanged. The same would happen if a lot of players over 1000 (the starting score) gets inactive, they will retain more and more points out of the pool of points getting exchanged.

So saying that a player with 1900 points is stronger than one that had 1600 points some weeks ago is likely a wrong assertion.

See also: viewtopic.php?f=5&t=9&p=5225#p5225
- viewtopic.php?f=7&t=493
- viewtopic.php?f=7&t=414

Comparing historical scores
One could say "ok, so how can I compare players according to historical scores?".

Well if we want to be pedantic, one cannot compare scores because those were achieved in certain conditions that are not repeated. One should provide the same opponents, the same level of skill for every opponent, the same type of AI controls, and so on. Too complicated.

But then could we approximate and say that a 1600 score now is equal to a 1600 score in 3 week in the future or 3 weeks in the past? No. It depends, accepting approximation, to the rest of the scores assuming that the score formula does not change [1]. If a 1600 score was in the past (3 weeks ago) a score held for a couple of games in the top league by the top player, while in the future (3 weeks from now) a 1600 score is the score normally held by players between the position 10 and 15, surely a score of 1600 was way more prestigious in the past than in the future. So relatively to the opponents active at that time, a 1600 in the past is way more stronger than a 1600 in the future.

But then we cannot compare at all? No we can still compare, but not at glance. Thanks to the score formula (see [1]) one could compute what a player could get with a certain score against his opponents at that time.

For example if in november 2016 a 1600 score player in the top league, winning the other players in the top20 could have get on average between 8 and 5 points, it means that the distance between the players was quite large and so the player was dominating. Instead if in january 2017 a player with 1600 score in the top league would get between 10 and 25 points from the top20, it means that the player is not weak but neither dominating. Actually he has often many over than him.

The problem is to get a "snapshot" of the ranking at a certain time (I'm doing this on telegram with screenshots), since mostly one can see the historical elo of one players, but not how strong was this score compared to the other players at that time.

Example from the telegram chat
###
Pier A:
TCT around 2200 is outstanding
NullPointer:
you're right that the ELO is fine now that I stopped to think about it.

He's ~8% ahead of the 2nd and is not losing many points with a loss, which is correct if look at the past.

Not that his score is not outstanding, but I was twice as higher than the 2nd (14%) back when I was playing hours in a row, which would be the equivalent of 2500 today and each loss would take between 20-30 (more than 1% of all my points) points from me, while today he loses 0.5%-0.8%.

So yes, I think the ELO is working fine
Pier A:
Hmm two flaws of your argument if I'm not mistaken. First you talk about percentages, that are wrong because the score is limited by matchmaking and elo formula. So you should account in terms of 'percentage of predicted success by the score formula' if you want to compare apples to apples. Percentages makes sense in other context. Again a confusion with the score model.

If you have 200 points and the second has 100,you have 100% more points, but the predicted winning percentage with the score formula is lower compared to someone that has 2200 and the second is 2000. Because what matters is the delta not the percentage. This is a confusion in using percentages. If I have 0 euro and I get 1, my improvement is outstanding in percentage terms, even compared to the one that had 1000 euro and ended with 5 gazillion.

As a reminder: smaller numbers, same delta, bigger percentage. Higher numbers, same delta, smaller percentage. Due to the score model, what counts is the delta.

Said that, tct has an outstanding score.

Second it depends how active were you at that time. When was that?
###

Further explanation: viewtopic.php?f=7&t=472

[1] the score formula should stays the same (because it can be tweaked and was already changed, see kfactor, halving points after draws [source on the old forum], etc.).
Last edited by pier4r on 18 Mar 2017, 10:16, edited 9 times in total.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 2768

Re: FAQ

Post#22 » 22 Jan 2017, 20:28

With the Xp system of alpha 7.8 , one could use the XP value and the number of victories to determine the average opponent score that was beaten by the player (percentiles would be better but in this case the average should be really close to the median).

Some examples:

xp = 1873887, victories=1345, avg opponent score=1393
xp = 1641018, victories=1187, avg opponent score=1382
xp = 3890994, victories=2815, avg opponent score=1382
xp = 3568450, victories=2572, avg opponent score=1387
xp = 362952, victories=245, avg opponent score=1481

So the last player seems to have gone through a way more difficult path (or not, it depends also on the score inflation).
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 2768

decorators

Post#23 » 29 Jan 2017, 23:09

See more about tagging: viewtopic.php?p=2161#p2187

Decorators
With bot tags (1) one can trigger different behaviors.

What is a decorator? Well the idea is from the decorator pattern (2). One problem of the game is that due to its competitiveness people try to optimize AIs for bot classes and maps, and that is so much work that even top players complains. (3)

Since I'm lazy, I'm ok to optimize per bot. So I have one main program for every bot, and slowly I build refinements through multiplayer testing (4). So The idea is to have basic AIs per bot that can handle well most of the fights given proper starting conditions. And how do I achieve those in particular maps?

Decorators!

Simple programs that tell the bot what to do at the start of a map. Using bot tags, as mentioned above.

Example of my AI that handle the initial conditions on mind game (I may discover further flaws while playing, it is just an example) and then calls the basic AI for the shotgun class.
Gladiabots_20170129_161522.png
Gladiabots_20170129_161522.png (222.76 KiB) Viewed 658 times


I hope this will help.

(1) that can be assigned to self or static objects like bases, that persist for the entire life of the self bot.
(2) http://wiki.c2.com/?DecoratorPattern ; Gladiabots, great for appreciating a lot of programming concepts and not only!
(3) here: viewtopic.php?f=2&t=379
(4) viewtopic.php?f=7&t=224#p3772
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 2768

elo and comparing two tactics

Post#24 » 06 Feb 2017, 10:46

A quick note about the score formula based on elo formula and comparing two tactics.

Reference: Elo formula

It is important to note that if a player has better tactics than another player, even for a tiny bit (like always winning only at 00:01 seconds left), if only those two competes, the score formula based on elo won't tell us the difference between those two tactics. This because the winning player will just keep accumulating points. The difference of points will grow more and more until cannot grow larger. So it will appear as one player is decisively stronger than the other, even if the difference on field is subtle.

Therefore the elo, on the long run, tell us how a player is doing against the other players in general. If the player if a tiny bit better than the others, he will have a way bigger score. Therefore the elo between players cannot be really compared but what can be compared is the relative position. If a player has a score higher than another, and both are pretty active, it is likely that the player with higher score has likely better tactics than the other player, even if for very minimal details.

But we cannot say that a player is 30% better than another. Just "either is better, equal, or worse", there is no way to quantify "how much" is better. Unless for direct fights, where the elo gives a percentage of the expected success.

Moral of the story: don't get discouraged if someone has an higher score than you, maybe you have just to apply tiny changes to get over him.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 2768

Re: FAQ

Post#25 » 23 May 2017, 12:15

reworked a bit the first post.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

Bocgwei1
Hello World
Hello World
Posts: 3

Re: FAQ

Post#26 » 16 Aug 2017, 16:35

Is their anyway to recover profile once the app has been Uninstalled then reinstalled. . It keeps saying profile already exists. ...

User avatar
Kanishka
Skynet
Skynet
Posts: 1404
Contact:

Re: FAQ

Post#27 » 16 Aug 2017, 16:43

Bocgwei1 wrote:Is their anyway to recover profile once the app has been Uninstalled then reinstalled. . It keeps saying profile already exists. ...


Make a new profile, and ask GFX to sync it.
Fixes break an AI more than bugs do. :ugeek:

Gladiabots Off-Topic Chat


My Stats: Kanishka_RN3;Kanishka_MiPad

Bocgwei1
Hello World
Hello World
Posts: 3

Re: FAQ

Post#28 » 17 Aug 2017, 05:40

Kanishka wrote:
Bocgwei1 wrote:Is their anyway to recover profile once the app has been Uninstalled then reinstalled. . It keeps saying profile already exists. ...


Make a new profile, and ask GFX to sync it.

Thanks

User avatar
Kanishka
Skynet
Skynet
Posts: 1404
Contact:

Re: FAQ

Post#29 » 25 Aug 2017, 07:54

Worth mentioning in the wiki: a tick is .25 seconds at normal speed.
Fixes break an AI more than bugs do. :ugeek:

Gladiabots Off-Topic Chat


My Stats: Kanishka_RN3;Kanishka_MiPad

User avatar
GFX47
Dev Team
Posts: 2090

Re: FAQ

Post#30 » 05 Oct 2017, 21:07

Maybe this could be moved to the wiki?

pier4r
Skynet
Skynet
Posts: 2768

Re: FAQ

Post#31 » 05 Oct 2017, 21:33

yes already planned.

WARNING: don't delete/hidden it. Taking away the sticky part is ok.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

Return to “Manual”

Who is online

Users browsing this forum: No registered users and 1 guest