ELO configuration

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
User avatar
Tisaac
Posts: 2327
Joined: 26 August 2014, 21:28

ELO configuration

Post by Tisaac »

Hi all,

The current version of the crew always grants the same thing at the end of a game.
This can be changed to mimic what is done in Hanabi : when a game end, we can setup the "opponent" ELO according to what happend during the game.
For instance, in 5p mode, making a perfect 25 is the same as winning against a 2090 ELO player.

What do you think should be the ELO computation for opponent in the Crew ? I will implement this in the new version so I need your help to configure this just right :)
User avatar
RicardoRix
Posts: 2106
Joined: 29 April 2012, 23:43

Re: ELO configuration

Post by RicardoRix »

yeah, don't make ELO a thing for co-op games - have you seen the division it creates?
User avatar
SpreadBoardGameJoy
Posts: 3
Joined: 17 February 2017, 23:13

Re: ELO configuration

Post by SpreadBoardGameJoy »

maybe a base elo of 3 per mission completed
- 1 per attempt after the first one
+ 1 if 5 players
+ 1 if later mission (maybe 30+?)
User avatar
Tisaac
Posts: 2327
Joined: 26 August 2014, 21:28

Re: ELO configuration

Post by Tisaac »

RicardoRix wrote: 18 January 2021, 17:04 yeah, don't make ELO a thing for co-op games - have you seen the division it creates?
I don't think Hanabi and The Crew are that similar, but this thread is precisely to discuss this topic.
huehnerlady
Posts: 13
Joined: 19 March 2020, 22:17

Re: ELO configuration

Post by huehnerlady »

I think discouraging leaving after 1 game would be good.

I am not sure with the specifics, but I would say should involve
* how much ELO you already have,
* how many players (therefore how difficult),
* how many rounds
* Which levels
* How many attempts
User avatar
8Barbara
Posts: 3
Joined: 05 April 2020, 13:09

Re: ELO configuration

Post by 8Barbara »

1. I think what's absolutely necessary to take into account:
- number of missions played
- difficulty of missions played
- number of fails to get there
2. I think it's absolutely necessary that you can also lose ELO. Otherwise ELO is just a sign of how much you've played.
3. It would be helpful, if it was a little easier to become average player, good player, strong player etc. Right now it is just about ok to set game restrictions to average players and above but sometimes you wait for ages... Still don't make it too easy either - e.g. that people are average players after 3 short games.
4. One central question will be, if you start ELO all over again or if you let people keep the ELO they have right now. If you have this essential part of people being able to lose ELO, I recommend people keep their ELO on the crew. People just playing to gain ELO but not being very good is a low ELO problem (difference between 5 and 50 most of the time).
User avatar
PurpleSunday
Posts: 1
Joined: 09 April 2020, 21:38

Re: ELO configuration

Post by PurpleSunday »

I understand the need to improve the ELO calculations. Right now it's a sign of experience, not skill, and has many flaws that have already been pointed out. But it will be hard to find an ideal formula for individual ELO in a team game. Good players can lose points with poor play from weak players. Weak players can get inflated ELO from good players who save them. Worst of all, Crew could go the way of Hanabi and be inundated with players demanding to abandon after one wrong move. I like playing Hanabi, but never really do so because of all the ELO manipulation there.

TLDR = I don't like Crew ELO, but I don't want a "fixed" ELO to ruin the experience either.
User avatar
LaViajera
Posts: 1
Joined: 29 October 2020, 18:39

Re: ELO configuration

Post by LaViajera »

ELO scoring on cooperative games is tricky business.
The rating should reflect the complexity of the mission, the number of missions completed, number of attempts needed and possibly also whether or not distress was used. I can see why it might be useful to include negative effects, however there is a lot of risk involved:
  1. Experienced players might be even less inclined to play with newbies, knowing that this may affect their scoring
  2. If you choose to leave a game because someone else is playing poorly, you might be the one punished for that (and not the poor player)
  3. There will be a lot more players requesting to abandon the game if they know there is a risk their ELO is going down after a bad play
User avatar
ufm
Posts: 1327
Joined: 06 January 2017, 08:38

Re: ELO configuration

Post by ufm »

Current ELO is just a mess. It does not reflect either experience or skill properly.

There are two major issues regarding this topic:

1. First-try quit

This is an ongoing one.
Players quit immediately after a single try to boost ELO, because the reward is always same as 1 point.
The Crew usually goes with 4 players, however a single player can stop the game right away.
If this happens, remaining participants have to find other players or tables repeatedly. What a waste of time and effort.

2. Game abandonment

This is widespread in Hanabi, and occurs in Forbidden Island a bit too.
When achieving perfect score becomes impossible, then the players collectively abandon the game to prevent ELO loss.
The practice is so common, and I once thought this to be BGA norm (actually it was not).
A few players even try to coerce others into abandonment.
And some players actively oppose abandonment, willing to bomb the game with abandonment players.
This is 'the division' RicardoRix mentioned.


So, what should The Crew's ELO do then?

As others suggested, ELO change should be weighted to the difficulty and the number of missions to discourage first-try quit. This is the basis.

Hanabi style is preferred to focus on measuring player skill, however this might cause the whole abandonment conflict for The Crew too.
Fortunately abandoning a Crew game is much harder compared to Hanabi though.
Abandonment players usually go with 2p games, because boosting ELO and abandoning are both easier than 3+ games.
Don't forget only a single vote can prevent abandonment!

The easiest way to shut this can of worms is, never reducing ELO in any case. Absolutely 0 advantage for abandoning games.
But in this case ELO will look like XP meter rather than a skill indicator.

Which one is the way to go? Honestly I can't figure it out. Both approaches have own pros and cons.
Maybe we could create a poll or something for this?
User avatar
marvel2010
Posts: 1
Joined: 15 June 2015, 01:03

Re: ELO configuration

Post by marvel2010 »

Tisaac wrote: 18 January 2021, 16:55 This can be changed to mimic what is done in Hanabi : when a game end, we can setup the "opponent" ELO according to what happend during the game.
For instance, in 5p mode, making a perfect 25 is the same as winning against a 2090 ELO player.
What are the options available when computing ELO? Is it possible to assign each player a different fake opponent that they won or tied against? For example, if there are players with ELO 200, 100, 50, can the player with 200 ELO win against a fake opponent with 2090 ELO and the 100 ELO player against a fake opponent with 1800 ELO?
Post Reply

Return to “The Crew: The Quest for Planet Nine”