@Farbs - Since this is your baby I'm tagging you right off the bat. I should note, I think the idea of your system is great, it just seems like some tweaking would make it a little less punishing. Recently, and I think greatly due to Flaxative's great work with the PvP Season, standing is getting a lot more talk. Some starting here, some in my post here. Based off the discussion on the PvP thread, I decided to do some research. Well I grabbed 30 games each from about 25 random players (just put in random names to neoncat's calculator), after filtering out all games that ended before either side scored 4 points (to eliminate quits) ended up with 635 games. Out of those 208 ended in a blitz or about 33%. Now to be clear only one player is getting blitzed in the game, so that would only be about a 16.4% blitzed rate per player. At that rate, the expected value for standing per a player who never quits is +.5087 per game. If the player is insta-quitting every other game to keep a low standing that would put them with an expected value of -1.4913 per every pair of games. Now if I understand the motivation behind standing correctly, it was to give +standing to players who don't quit, and negative standing to players that do. But in my personal opinion it seems like it should be weighted evenly on both sides of zero. Below are some somewhat arbitrary expected values if different values were used. Feel free to discuss. Rows in blue are the theoretical standing +/- values, left column gives expected value per game of someone who never quits, right column the expected value for the every other game quitter for each pair of games (one normal, one quit).
This is awesome, thanks for doing this study and I hope it helps @Farbs improved standing. I like that blitzing (in all forms) matters more for guild combat, but I definitely agree that getting blitzed while being "blameless" (no premature quitting) shouldn't hurt a player's standing. Especially when you see honest players like @gulo gulo essentially quitting ranked play because their standing treadmill is just too brutal, or players like you, @Scarponi, who are trying their best to make amends for past tanking.
Between +1 and -2 there are 0 and -1. Those amounts of standing could be awarded for less tanking-like behaviour, but the risk is that tankers are encouraged to behave in more annoying ways.
It might help if number of rounds played was available as a modifier. It's possible for someone to take 5-8 rounds to throw a game, but unlikely IMO.
That would definitely be another viable option. It seems if a change was to be made to adjust for blitzing, one could either change the scoring system of Standing, or change the parameters of the scoring system. Therefore, if the +1/-2 is preferred, as you say, a set number for how many rounds have been played could be put in to eliminate a blitz, or the victory point marks could even be moved (only having to score 4-1 instead of 4-2, or even 5-1 could potentially be better). I would have to crunch some more numbers to see how things like that would affect Standing's expected value. Ultimately I think there's two questions at play: 1) How many games do you want someone to have to play to bring their standing back to even after a quit (whether insta-quit, rage quit, lost connection, child that needs immediate attention, etc.)? As it currently stands based on the above data it would take 4. [Expected value for clean play is +.5, the quit will cost the player -2, so you have -2 + 4(.5)=0.] Personally I think that's too much. 2) Is there a way to accommodate playing styles/teams that have a higher likelihood of being blitzed? The average rate of being blitzed is 16.4%, that means some players are getting blitzed at a higher percentage (resulting in a lower expected value). Is there something to be done for the super aggressive player that might get blitzed 25% of the time and would have an expected value of .25 or about half that of the average player?
While I have been trying to get back into the game after quelling my frustration in relation to Standing (ultimately it doesn't mean anything unless someone is going to utilize it for something), I appreciate the thoughts people have put down. I think the Standing system is a good system, it does need to be tweaked. However, I don't think this is something that should not come into consideration that much when making any changes, even if it would affect my own personal standing greatly. Part of the game is the risk and reward of the choices you make in the game. You may have a combination of things that gives you a chance of blitzing a lot of players, but you also will be blitzed often as well? That's part of the game, and you have to accept it. That being said, players playing this style should expect to never be at the top of the Standing leader boards. However, they also should not be penalized by the same system for playing within the rules. One of the things Standing does not do is recognize when you blitz someone else. If being demolished has such a negative effect, why wouldn't the opposite also hold true? Just a thought.
Having wins and blitz wins earn more standing than non-blitz losses would enable some higher risk strategies. Just have to make sure it's not so generous that tankers can recover their standing on their up-streak. Another thing standing could take account of is relative rating. You're more likely to get blitzed by someone with much higher rating than you, so it's less likely that such blitzes are a result of tanking and so deserve less punishment than blitz losses against lower-ranked players. If standing is sensitive to the rating lost in a game it can more effectively punish the act of tanking while minimising the harm to honest players.
