Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

This is the home for suggestions for site improvements, changes to house rules, and new variants.
Forum rules
It's okay to suggest new rules variants in this forum, but proposing new *maps* should be done in the linked "New Map Variant Proposals & Voting" subforum.

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby NoPunIn10Did » 05 Jun 2018, 18:07

NJLonghorn wrote:
This. When a game ends 15-15-4, the player who managed to cling to life deserves full credit for the draw. My favorite game of all time was when I survived as the 1 (Tunis) in a 17-16-1 draw. It took all I was worth to convince both other players that they couldn't trust the other to try to get rid of me. Downgrading my credit for that game would be a HUGE step in the wrong direction.
Giving the 15s/16s/17s extra points would give them an extra incentive to accept a draw rather than try to finish off weaker players.

Several things to unpack here:

  1. If you really like maintaining equal-shares-of-the-draw, that won't be going away.
    • You could still play games using that option.
    • I anticipate that ranked draw games will have a sizable niche following and won't be the majority of games played.
  2. The scenario you're describing (the 17-16-1) isn't actually comparable in a rank-based draw scenario.
    • It would occur under a very different negotiation framework; it's important not to assume that games under the DSS system would play out that way in the first place.
    • While in a Draw-Sized system, eliminating opponents is the only way to increase one's final draw score, eliminating opponents in a rank-based system has no value in-and-of-itself.
    • What this tends to mean is that the larger powers will be fighting more among themselves as they jockey for position, rather than spending all their efforts on wiping on the small fries.
    • In a rank-based game, the smaller powers have more options than simply to argue for their continued existence. There's a smoother continuum between final game results, so they can argue for more varying degrees of success rather than a simple binary of eliminated/included.
    • There's more room to grow back from the brink: a small power, since it's less likely to be under direct threat by the big powers, can snatch SCs from comparably sized opponents.
    • Even if you are eliminated, you're rewarded for holding out longer than other eliminated opponents.
  3. There are several years of experience in the F2F tournament circuits that have shown that it's not a "HUGE step in the wrong direction."
    • It definitely changes the game, but that's intended.
    • I don't believe there's any tournament that still uses 100% equal draws.
    • Modern tournaments use a variety of systems, but with the exception of Dixiecon (an odd hybrid of draw size and SC count), they've all moved to center-based, lead-based, or rank-based draw calculations.
    • Even the PDET had a lead-based tiebreaker component (SOS).
    • webDiplomacy, one of our most popular competitors, offers multiple scoring options, including Sum-of-Squares (which is lead-based rather than rank-based, but largely similar in play). It hasn't collapsed.
    • The point of all this is to say that what's being suggested here is simply another mathematical route down a well-worn path.
    • Rank-based and lead-based scoring have been around for years. They haven't ruined the game.
  4. As to solos:
    If you want to incentivize solos, award extra points for solos. The game is too dynamic to do it any other way.

    • That's still going to be the case; I'm not suggesting a change to how solos are scored.
    • What actually encourages solos within rank-based systems is all the infighting it promotes.
    • In draw-sized scoring, there's little reward to betrayals within an Anti-Leader-Alliance, since you have to completely wipe out an opponent to make such a betrayal worthwhile scorewise.
    • In rank-based scoring, there's more room for the players in an ALA to engage in shady tactics as they jockey for position, since even a modest betrayal can lead to considerably more points.
    • In practice, this can make anti-leader-alliances more fragile, providing more opportunities for a skilled player in the lead to find a crack in the walls.
    • Games with more opponents still on the map are those that are more likely to end in a solo. The more opponents are removed (particularly early on), the more likely a game will end in a draw. Scoring systems that encourage attacks on the leaders, rather than directing fire to the smaller powers, tend toward fewer eliminations overall.
Lead Volunteer Developer & Forum Admin

Variant GM & Designer
User avatar
NoPunIn10Did
 
Posts: 2748
Joined: 17 Aug 2011, 00:17
Location: North Carolina
Class: Ambassador
Standard rating: (1000)
All-game rating: (1466)
Timezone: GMT-5

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby jay65536 » 05 Jun 2018, 21:07

So, on this thread, I'd say 2 things.

1) I'm not sure this is an appropriate thread to discuss the different merits of rank- vs. draw-based rating systems. (And furthermore, I think there's a big difference between SCORING systems for tournaments, and RATINGS systems for online "long games".) All I'll say here is what I'm saying below--if you want to debate me, there are other threads open where I'll do that. My opinion as it pertains to this thread is that if you decide to introduce a purely rank-based scoring option, I believe this system is a good one compared to other rank-based systems that could be tried instead.

2) Arguing, using anecdotal evidence from individual games, about which rating system is preferred is, to me, nearly pointless. The reason is because BOTH draw- and rank-based ratings are going to have "edge cases" that are basically results that clash with what you want to reward in the system. NJLonghorn has given an edge case for rank-based, but if you want an edge case for draw-based scoring, you can start with literally any game in which the players voted a 2way draw before they actually made it to 17/17. (There are others too.)

For each story purporting to show a flaw in rank-based ratings, there are others that show flaws in draw-based ratings.

Ultimately, the point of having a rating system is to measure skill; different systems measure different skills. And one would hope that since these games are not being played for money or any other value than fun, if you want to play the game a certain way then you will regardless of rating system (or, as NP said, you'd just avoid the games that are rated in the way you don't like).
jay65536
 
Posts: 482
Joined: 10 Sep 2016, 18:13
Class: Ambassador
Standard rating: (1120)
All-game rating: (1126)
Timezone: GMT-5

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby asudevil » 05 Jun 2018, 22:19

NJLonghorn wrote:
LordJezza wrote:Hmm, I guess my main point was that i don’t think it is true to say just because someone has more SCs, then they are closer to a solo... I understand however, that this is a matter of perspective. Definitely if this encourages more solos, I’d be for it, but at its basest level (as I understand it) is punishes those who scrap for a draw and make themselves so indespensible, that a draw cannot happen without them. This is diplomacy after all, and if a player with few SCs is able to convince others that they can’t possibly eliminate him, then I see that equally as a person forced to share in a draw despite being on 17.


This. When a game ends 15-15-4, the player who managed to cling to life deserves full credit for the draw. My favorite game of all time was when I survived as the 1 (Tunis) in a 17-16-1 draw. It took all I was worth to convince both other players that they couldn't trust the other to try to get rid of me. Downgrading my credit for that game would be a HUGE step in the wrong direction.

Plus, I don't buy the idea that the OP proposal would encourage solos. Giving the 15s/16s/17s extra points would give them an extra incentive to accept a draw rather than try to finish off weaker players.

If you want to incentivize solos, award extra points for solos. The game is too dynamic to do it any other way.


This is my issue every time this comes up. If you can wedge yourself into a draw that they CANT eliminate you...that's awesome. IMO (player not mod) rank based draws encourage solos...only because it makes it harder to STOP the solo. People are more likely to "pick off" an SC from an ally to raise their draw score...which destabilizes the ALA. ALA's are hard enough to put together an make work. The guy who already is the smaller one may not be able to gain SC...but he is vital in setting up the line. He may even be able to lose a couple SC and someone in the rear of the stalemate line could pick off that guys SCs...hurting his score. So when you are the 5th best score in a draw...why not just lose to a solo if someone decides to start picking off SC.

I mean hell, you basically saw this happen in the TTT. 1812 kinda ruined the entire tournament because USA was trying to pick off SCs to raise their rank in SCs for the draw...and Britain/Russia said fuck it...we will just lose to a solo. This same mentality and "kingmaking" (which people already think is an issue) will only become more prevalent under rank based draws.

That's my thoughts. Again, as a player. But unless you are the STRONG second or third...why would you fight hard for an ALA as a 2/3 SC 4th or 5th power...instead of just letting them solo since your point differential between them would be so small...and people are going to be more concerned with WHO is getting the SCs as you set up a stalemate line...since its rank based...making ALAs harder.

I strongly dislike this plan.
Captain FANG, forum team championships WINNER
Part of the surviving nations of WW4/Haven

Unless I am in the cheater's subforum. 99% of what I say is NOT as a mod.
User avatar
asudevil
Premium Member
 
Posts: 16599
Joined: 18 Jul 2011, 02:20
Class: Star Ambassador
Standard rating: (1351)
All-game rating: (1437)
Timezone: GMT-7

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby asudevil » 05 Jun 2018, 22:24

NoPunIn10Did wrote:[*]The scenario you're describing (the 17-16-1) isn't actually comparable in a rank-based draw scenario.


But we don't have DIAS as all draws...so people can ask for the draw long before this point of the game.


[*]There's more room to grow back from the brink: a small power, since it's less likely to be under direct threat by the big powers, can snatch SCs from comparably sized opponents.


Therefore decreasing the success of ALA


[*]I don't believe there's any tournament that still uses 100% equal draws.


Tournaments also only are running 3 rounds or so...so they have to differentiate a LOT more than when we have hundreds of games in our rating.


[*]Even the PDET had a lead-based tiebreaker component (SOS).


Again, smaller sample size

[*]webDiplomacy, one of our most popular competitors, offers multiple scoring options, including Sum-of-Squares (which is lead-based rather than rank-based, but largely similar in play). It hasn't collapsed.


They suck...so I don't care what they do. Their entire interface and forum is a disaster.


[*]What actually encourages solos within rank-based systems is all the infighting it promotes.
[*]In draw-sized scoring, there's little reward to betrayals within an Anti-Leader-Alliance, since you have to completely wipe out an opponent to make such a betrayal worthwhile scorewise.
[*]In rank-based scoring, there's more room for the players in an ALA to engage in shady tactics as they jockey for position, since even a modest betrayal can lead to considerably more points.
[*]In practice, this can make anti-leader-alliances more fragile, providing more opportunities for a skilled player in the lead to find a crack in the walls.


Its already SO hard to do a successful ALA in an online forum ... making those more fragile doesn't really help the situation...again...IMO
Captain FANG, forum team championships WINNER
Part of the surviving nations of WW4/Haven

Unless I am in the cheater's subforum. 99% of what I say is NOT as a mod.
User avatar
asudevil
Premium Member
 
Posts: 16599
Joined: 18 Jul 2011, 02:20
Class: Star Ambassador
Standard rating: (1351)
All-game rating: (1437)
Timezone: GMT-7

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby asudevil » 05 Jun 2018, 22:25

jay65536 wrote:but if you want an edge case for draw-based scoring, you can start with literally any game in which the players voted a 2way draw before they actually made it to 17/17. (There are others too.)


In which case we should just make all rated games...DIAS...that would encourage soloing...because its VERY hard to get that "2man draw" to the 17/17 line...where its much easier to get to like 12/14 and just convince everyone else to give it to you.
Captain FANG, forum team championships WINNER
Part of the surviving nations of WW4/Haven

Unless I am in the cheater's subforum. 99% of what I say is NOT as a mod.
User avatar
asudevil
Premium Member
 
Posts: 16599
Joined: 18 Jul 2011, 02:20
Class: Star Ambassador
Standard rating: (1351)
All-game rating: (1437)
Timezone: GMT-7

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby DQ » 05 Jun 2018, 22:42

jay65536 wrote:2) Arguing, using anecdotal evidence from individual games, about which rating system is preferred is, to me, nearly pointless. The reason is because BOTH draw- and rank-based ratings are going to have "edge cases" that are basically results that clash with what you want to reward in the system.


This! I'm not a long-time player on the site, but having played in F2F tournaments for 20+ years, I can say two things:

1. The way the games are scored definitely affects play, and
2. This only ever matters in edge cases.

ANY scoring system (ok almost any scoring system) will result in the "better" players getting "better" results 95%+ of the time. So the question is ... how much are you fighting for the 5%?

For my money, the only risk you run by adding Fibo-Dip as an alternate scoring method is having someone game the system to increase their ELO - which ... is currently done by draw whittling, which is massively unpleasant. Would the meta-gaming under Fibo-Dip be more unpleasant? I haven't found it to be the case in F2F games! In general, much less angst about the delta between coming in second or third - but there IS still concern about these things!

I'm in favor.

DQ
Stab you soon!
User avatar
DQ
 
Posts: 355
Joined: 11 Sep 2008, 14:29
Class: Diplomat
Standard rating: (1062)
All-game rating: (1063)
Timezone: GMT-5

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby DQ » 05 Jun 2018, 22:48

asudevil wrote:
jay65536 wrote:but if you want an edge case for draw-based scoring, you can start with literally any game in which the players voted a 2way draw before they actually made it to 17/17. (There are others too.)


In which case we should just make all rated games...DIAS...that would encourage soloing...because its VERY hard to get that "2man draw" to the 17/17 line...where its much easier to get to like 12/14 and just convince everyone else to give it to you.


I'm trying to understand your point here - you're arguing against an alternative scoring structure option because you don't agree with the incentives that system provides, yes?

IF that is the case, then ... could you explain what you think the harm of having an alternative scoring system would be, for the site and upon the ranking system?

IF not, can you help me understand your point? I don't disagree with the main thrust of what I see to be your argument ABOUT the scoring system - which I read as that it incentivises / rewards different kinds of behaviors - but given that no one is asking you to play under that system, this doesn't seem relevant?

NoPun has outlined some reasons why he thinks having an option would be good for the site, and would like to see the method (which has worked IRL pretty well over the last 10+ years) given a chance to succeed/fail on its merits here. Is that something which you oppose?
Stab you soon!
User avatar
DQ
 
Posts: 355
Joined: 11 Sep 2008, 14:29
Class: Diplomat
Standard rating: (1062)
All-game rating: (1063)
Timezone: GMT-5

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby V » 06 Jun 2018, 02:49

I’ve already commented my lack of interest in scoring systems in general, but can see two points possibly worth mentioning.

1. It’s being proposed as just an option, that folks like me are welcome to avoid, so no harm done. Unless it’s an onerous amount of work for the site administrators for possibly little use, then it should get the appropriate low priority.

2. It appears even the advocates for the change point out that it significantly alters the way the game is played & in fact for some that is the objective. At that point should it be considered “standard”, or would it have to fall into the “unranked” (like solo only games are currently) or “all games”, in amongst the variants?

We’ve already had an issue with folks playing specific game variants that boosted their “site rating”, then complaining when the new scoring system flattered them a little less. I doubt we want a rerun, following folks realising “aha, let’s click this option, we get more points that way”.

I like simple systems & level playing fields. My maths is not up to establishing if this idea poses a threat to what is right now an apparently excellent system currently in use. I leave it to others to make that assessment, but hope they do before introducing something that causes more problems than benefits.
Platinum Classicist
Voilà! In view, a humble vaudevillian veteran cast vicariously as both victim and villain by the vicissitudes of Fate. This visage, no mere veneer of vanity, is a vestige of the vox populi, now vacant, vanished.
V
 
Posts: 739
Joined: 04 May 2014, 21:28
Class: Star Ambassador
Standard rating: 1778
All-game rating: 1829
Timezone: GMT-7

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby NoPunIn10Did » 06 Jun 2018, 02:56

V wrote:We’ve already had an issue with folks playing specific game variants that boosted their “site rating”, then complaining when the new scoring system flattered them a little less. I doubt we want a rerun, following folks realising “aha, let’s click this option, we get more points that way”.

I like simple systems & level playing fields. My maths is not up to establishing if this idea poses a threat to what is right now an apparently excellent system currently in use. I leave it to others to make that assessment, but hope they do before introducing something that causes more problems than benefits.


I can assure you that this is not going to be any easier to use for score-boosting; that's actually part of the mathematics I've walked through. Any of the strategies required to metagame this particular scoring option are the same as what are already periodically used by players who try to cheat the site, and the points awarded to a given player are still quite close to the points they'd already receive in a draw under the current system. It won't be the most efficient way to "game the system," so to speak.
Lead Volunteer Developer & Forum Admin

Variant GM & Designer
User avatar
NoPunIn10Did
 
Posts: 2748
Joined: 17 Aug 2011, 00:17
Location: North Carolina
Class: Ambassador
Standard rating: (1000)
All-game rating: (1466)
Timezone: GMT-5

Re: Ranked Draws & Rank-Based Scoring: Fibonacci-Diplo

Postby super_dipsy » 06 Jun 2018, 07:19

V wrote:1. It’s being proposed as just an option, that folks like me are welcome to avoid, so no harm done. Unless it’s an onerous amount of work for the site administrators for possibly little use, then it should get the appropriate low priority.

Unless I have misunderstood the suggestion, this system would work in conjunction with the current one. In other words, playing games with this option will affect your site rating too, but just differently. The only way to avoid this would be to say that games played with this option are unranked in terms of our current structure, but I don't think that is what is being proposed.

Nopun, I saw your response about how you would handle surrenders and pick-ups, but it made me realize I don't quite understand the proposal. Can I play back what I think I now understand to be the way this works?

- We still use our elo-like current system
- For games with this option, it only takes effect if there is a draw (is this right? or is the solo result modified too?)
- In a draw, the way points are awarded will be altered; instead of the sharing of pointss done today, you have a different scale based on rankings which would be applied to the rating algorithm to calculate the deltas
- An important point which I am not clear on - am I right in thinking that with this system you could get points for being eliminated? Eg if I am the last to go in a 2-way draw, I get points because I was ranked third?
- Can I also get points if I surrender (ie I was last to surrender in a 2-way draw)? If not, then do you have to adjust the scoring allocation based on the fact tht only 2 players are going to participate in the rank-based scoring?
- If you surrender, you simply get what our current system would normally give you for a surrender
- If you pick up a country, you get a pro rata of your eventual result based on turns played as we do today

Have I got that right? I am particularly interested in the question about eliminated players getting points or not. Did I misunderstand that?
User avatar
super_dipsy
Premium Member
 
Posts: 12077
Joined: 04 Nov 2009, 17:43
Class: Ambassador
Standard rating: (1000)
All-game rating: (931)
Timezone: GMT

PreviousNext

Return to Suggestions

Who is online

Users browsing this forum: No registered users and 8 guests