A (Choose: Great/Pointless/Annoying) Scoring System Thread

Anything about the Diplomacy game in general.

A (Choose: Great/Pointless/Annoying) Scoring System Thread

Postby jay65536 » 22 Nov 2017, 06:31

I was having a discussion about scoring systems with some people on a different thread, and I realized the thread was being hijacked, so I figured I would just start a new thread. I haven't been able to find a thread like this on the forum so far, just dedicated to chronicling (and expressing opinions about) different scoring systems. In particular it seems from my short time here that many site regulars are extremely invested in draw-based scoring, which has nowhere near that level of primacy among live tournament players.

There are many, many different scoring systems that have been invented for Diplomacy. All of them are meant to deal with the fact that when someone can't solo, you have to reward draws somehow. All scoring systems are fundamentally based around the same question. If you can't solo, then you should be rewarded more the "closer" to a solo you are; but what does it mean to be "close to a solo"?

Scoring systems--at least the ones I'm aware of--tend to be organized around 3 major answers to this question. There are also other questions that a scoring system is trying to answer, like: should all games of Diplomacy be worth the same, or should some be worth more than others? (Put another way: is Diplomacy a zero-sum, or at least a fixed-sum, game?) A sub-question of that is, how good does a string of non-solo results have to be to overtake a solo, and should it be within the realm of possibility, or should solos always take precedence?

The three major answers to the question of "closeness to a solo" are:
1. The fewer players in the draw, the closer you are to a solo
2. The more centers you have, the closer you are to a solo
3. The bigger your lead in centers is over your competition, the closer you are to a solo

In practice, most scoring systems are going to be some kind of hybrid of these 3 ideas.

The first idea represents draw-based scoring. In draw-based scoring, the next-best result to a solo is always a 2way draw, then a 3way draw, all the way down to a 7way draw. As an example, the "purest" such system is called Calhamer points. The Calhamer point system basically says that all games will be worth the same amount of total points--let's imagine it's 420 so all the numbers are integers. Then if someone solos, they get 420 points; in a draw, all draw participants equally divide the 420 points. So you'll either get 210, 140, 105, 84, 70, or 60. A loss gets 0.

I say this is the "purest" draw-based system because it is the only one that does both of the following two things: it treats Diplomacy as a fixed-sum game; and it adheres to the principle, laid out in the rules, that if the game does not end in a solo then all survivors share equally in a draw.

Of course, in practice no one uses Calhamer points. People tweak this system all the time, sometimes by just mixing in something else as a tiebreaker and sometimes changing it in bigger ways.

An example of a draw-based system that is used in real life is the Dixiecon system--it's primarily a draw-based system, with a center-based component as well (basically for breaking ties). Unfortunately, because I haven't been to Dixiecon in awhile, I don't remember the finer points of the system. (Maybe someone else reading this does?) The PDET system used this year is also primarily a draw-based system; it's actually very close to Calhamer points, but it made the changes that a solo was worth more total points (760) than the sum of any draw (600), and it used a tiebreaker (sum of squares) that is not draw-based. One thing that most draw-based systems have in common is that it is usually possible to put together a small (3-4) string of draws that can overtake a solo and a string of losses.

The second idea represents center-based scoring. In a center-based system, the only thing that matters is how many centers you have. Having 17 centers is always better than having 16, is always better than having 15, etc. Most of the time in tournaments I've played in, center counts are used as secondary components of other systems, to break ties and such.

I don't think I've ever played in a tournament that used center-based scoring primarily, but I think an example from real life is the GenCon system. It's very different from any system I've ever played under. Here's how it works. The last round contains a "top board". In order to get on the top board, you have to have the highest center count *with a specific country* over all of the previous rounds. So, the player with the highest center count as Austria gets in; so does the player with the highest center count as England, etc. (There are protocols for what happens in the event of ties, or if one player has more than one of these slots; but they all are based on the center counts, never on anything else.) Then, whoever finishes the top board with the most centers wins the tournament.

The third idea represents lead-based scoring. The premise of a lead-based system is that being first in the center count should always count as being "closer to a solo" than any other result. Sometimes the amount you're leading by is a factor; other times it's not. But the basic premise is simple: if you have 17 centers and someone else also has 17, you have not done as well as if you have 6 and no one else has more than 5.

Lead-based systems have become more common since I first started playing Diplomacy. The reason is that they reward aggressive play, which usually makes for more fun games. In a draw-based system, if one player pushes for the solo and gets stopped by a coalition of 4 powers, the game won't end there because everyone has an incentive to play on and whittle the draw down. Or, a coalition of powers has agreed from the start to play for a small draw, which is usually antithetical to playing to win. But in a lead-based system, the incentives are totally different. When the players are no longer playing to win, or at least not to lose, they have no incentive to play on.

A really simple example of lead-based scoring is what's called the Carnage system (named after the gaming con where it was invented). It's primarily lead-based, with a center-count component as well. It also totally throws away the idea that Diplomacy is a fixed-sum game. If the game ends in a draw, the board leader gets 7000 points plus their center count; 2nd place gets 6000 points plus their center count; all the way down to 1000 for 7th place. (Ties split the difference.) If the game ends in a solo, the winner gets 1,000,000 points--solo winners cannot be caught by any number of draws!

There's another common system that's a hybrid of lead- and center-based scoring, called Sum of Squares. At the end of each game, you square everyone's center counts and add up those numbers; your score is your center count squared, as a percentage of that total. The exception is when someone solos; then they get 100 points. This means each game is worth the same number of total points--100. It also rewards you handsomely for topping the board and building up a significant lead over your closest competitor. For example, if the final center counts are 13/10/6/5/0/0/0, the table-topper's score is going to be 100 * 169/(169+100+36+25), or about 51.21. Two of these results will actually beat a solo and an elimination in this system! This is the system that the PDET used as its tiebreaker, secondary to the draw-based component.

(C-Diplo is another really common system, but I've actually never played a C-Diplo tournament.)

So there are a lot of different ways of scoring out there. If you are only used to playing in draw-based systems, I'd say don't knock lead-based systems 'til you've tried 'em. If you are going to express a preference for a certain scoring system, you should know what else is out there too.
jay65536
 
Posts: 420
Joined: 10 Sep 2016, 18:13
Class: Ambassador
Standard rating: 1120
All-game rating: 1126
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby DQ » 22 Nov 2017, 17:03

jay65536 wrote:The three major answers to the question of "closeness to a solo" are:
1. The fewer players in the draw, the closer you are to a solo
2. The more centers you have, the closer you are to a solo
3. The bigger your lead in centers is over your competition, the closer you are to a solo
finishes the top board with the most centers wins the tournament.

So there are a lot of different ways of scoring out there. If you are only used to playing in draw-based systems, I'd say don't knock lead-based systems 'til you've tried 'em. If you are going to express a preference for a certain scoring system, you should know what else is out there too.


Someone was recently asking me about solo victories, and how many I've gotten over the time I've been playing (f2f) - Its at least a dozen, as those have been captured by WDD. Which I only mention to say that the theory behind point 1 is, in my experience simply false. The fewer players there are in the draw, the further you are from a solo, almost always. You want chaos, and people getting on each other's nerves, and tension, and ideally someone on the other side of the line who just doesn't care if you win or not. The more people there are over there, the better for you. If its just you on 14 and two other guys on 10, you really shouldn't get to 18. People do, of course, but that's another story.

I've seen plenty of people - world champion Doug Moore, recently - stopped on 17 by a coalition of 5 players. But he was definitely CLOSER to the win than any of the other players.

Lead based scoring (and I note for the record that Carnage uses a Tier System whereby Solo wins are superior to any combination of results) also tries to capture this - notably the "sum of squares" systems, which normalize all scores to 100 points for every board. Carnage is also zero sum. The idea has emerged that there is some merit to "topping" the board, and topping the board with greater separation suggests you are closer to winning (15/10 > 15/14).

Religious debates around scoring systems are eternal, and fun, because there is no authority and no one can tell you you're interpretation is wrong. /puts on NADF hat "We're working to change that! /takes off hat

At the end of the day, scoring systems almost always identify the correct winner. Good players will adjust their target towards the goal they are given, and the changes in scoring systems mostly affect players in the middle ranks. "Under Carnage I would have come in 11th instead of 23rd!" *cough* who cares? *cough*

I'm much, much less familiar with the online scoring system conventions. What do folks use for tournaments here?
Stab you soon!
User avatar
DQ
 
Posts: 315
Joined: 11 Sep 2008, 14:29
Class: Diplomat
Standard rating: (1062)
All-game rating: (1063)
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby super_dipsy » 22 Nov 2017, 18:14

Excellent summary of what is around and how they all work! Great job and very informative :)

I am interested in your statement that
jay65536 wrote:The reason is that they reward aggressive play, which usually makes for more fun games. In a draw-based system


I would be most interested in how you guage that and any stats you have. From many of the discussions here over the years, a common thrust has been that people want to play where more people go for a solo rather than the draw. Your name for the sort of system we have (draw-based) seems to imply that it will lead to more draws, whereas the comment about the lead-based system seems to suggest a bigger push for solos? (or is more aggressive a general comment rather than one on solos?).

My interest is because there almost seems to be a contradiction in what we see here on Playdip and what we might expect. We generally run at a higher overall percentage of solos than most other sites where I have been able to view the statistics, and also as judged against 'accepted' rates in postal versions of the hobby (seems like historically the average used to be a solo roughly 1 game in 3). Admittedly we do tweak the 'draws-based' approach by uplifting a solo considerably, so that a draw is worth a lot more than two 3-way draws, and it is also possible that because our system is Elo-like it can sometimes mean people know they have to solo to get anything on their ratings, but even so I am a little surprised our solo rate is as high as it is.

But perhaps the question of an 'aggressive' game does not equate to a solo rate. Maybe people get too hung up on solo rates when they are looking for a 'good game'. Perhaps the lead-based approach would encourage more general aggression but not necessarily more solos (or actually, I suppose he who has the most SCs gets a 'solo' anyway?). Hmm - that is another question I have. I am not sure I understood right. In the lead-based system, if I get to 18 centres, do I get a bigger reward than agreeing a draw with 16 centres? But certainly an intereting idea, and a good discussion. At first galnce though (and I may be misunderstanding), doesn't the lead based approach suffer from the same likelihood of an equivalent to draw whittling as in SC whittling? If someone is out in front with SCs, you may not want to agree a draw until you can team up with some others to drag them back a bit. Not sure it is the same, but it feels similar.
User avatar
super_dipsy
Premium Member
 
Posts: 12068
Joined: 04 Nov 2009, 17:43
Class: Ambassador
Standard rating: 1000
All-game rating: 931
Timezone: GMT

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby jay65536 » 22 Nov 2017, 23:27

DQ wrote:the theory behind point 1 is, in my experience simply false. The fewer players there are in the draw, the further you are from a solo, almost always. You want chaos, and people getting on each other's nerves, and tension, and ideally someone on the other side of the line who just doesn't care if you win or not. The more people there are over there, the better for you. If its just you on 14 and two other guys on 10, you really shouldn't get to 18. People do, of course, but that's another story.

...

I'm much, much less familiar with the online scoring system conventions. What do folks use for tournaments here?


DQ, to your first point, I agree completely; your experience matches mine 100%. I only included that in my OP because it is the philosophy underlying draw-based scoring. I was trying to play a bit of devil's advocate at certain points in my OP.

To your second point, I just wrapped up playing in the PDET, a tournament that's been running on this site for most of the year. The system is primarily draw-based, with a sum of squares component. The draw-based part is 760 for a solo, and 600/N for an Nway draw. Then all draw participants have their sum of squares score added in as a decimal (so not multiplied by 100). Losses always got 0 points. As an example, my last-round game ended with me on 4 in a 15/8/7/4 4way; my score from that game was 150.0452. I am currently under the impression that the draw-based system is universally embraced but the inclusion of a non-draw-based tiebreaker is controversial. This is based on two things: me reading in the forum that a previous tournament didn't have a tiebreaker and had to play extra rounds to break ties; and me seeing that when someone took over the tournament from Carebear they didn't bother to include the SoS component in the tournament leaderboard.

Also, Carnage is not zero-sum. It's just not. No scoring systems currently in use are zero-sum, and only a handful (like sum of squares) are fixed-sum. Carnage is neither.

super_dipsy wrote: From many of the discussions here over the years, a common thrust has been that people want to play where more people go for a solo rather than the draw. Your name for the sort of system we have (draw-based) seems to imply that it will lead to more draws, whereas the comment about the lead-based system seems to suggest a bigger push for solos?


It's called "draw-based" because it's scored based on the size of the draw, not necessarily because it encourages draws. Whether draw-based systems lead to more solos is not clear, although my experience points to the notion that it doesn't.

Here are two stories that could serve to illustrate either side of this argument. The first story is a house game where a player who was used to a draw-based system coughed up a solo in the endgame. It was my first FtF game with members of the traveling hobby. I was a 12-center Turkey, in a coalition with a 4-center Italy and a 2-center France. We had stalemated a 16-center England. However, Italy did not want to end the game in a 4way draw. He wanted to whittle France out and get a 16/12/6 3way. I told him he would be unable to do that, but he didn't care. He just didn't want a 4way. So England backed off the line, Italy stabbed France, and England made a well-timed run back to the line to win a solo that France threw him.

In any system besides draw-based, England would have just gotten stalemated, taken a 16/12/4/2 result, and not soloed. Playing the Carnage system, he would have gotten 7016, the second-highest possible score you can get for a draw; playing sum of squares, he would have gotten 61.24. But playing in a draw-based system where one of the players would not take a 4way under any circumstances, he soloed.

The other story is my most recent game--round 3, PDET. The game ended G15 / A8 / I7 / T4. There was one spring turn, when England, Russia, and France were all still alive, and Germany was clearly the board leader, where he had to decide: push for a solo? Or play passively, mop up E/F/R and let A/I kill me, and take a safe 3way draw? Germany made the decision that he had to at least try to solo the game. So he bypassed E/F and pushed his fleets toward the Med, while also pushing armies east. A/I immediately reacted by pulling off me (T), I ended up helping to set up the stalemate line, and the ending position was one where no one felt comfortable trying to whittle the draw down. As a result, Germany took a 4way.

The salient question here is, did Germany make a mistake pushing for the solo? If we're playing a draw-based system, there's a valid case that he did. The good thing about lead-based systems is that it is NEVER a mistake to push for a solo! In the PDET system, the German player is going to get a score of 150.6356. That's a worse score than any 3way. In sum of squares he'd get 63.56, a MONSTER score for a non-solo. And in Carnage he'd get 7015, the third-best draw result possible.

So I think there are certain times when either draw-based or lead-based scoring could lead to solos that wouldn't have happened under the other type of system. Draw-based systems can lead to solos being coughed up by people who are trying to whittle down draws and not doing a good enough job; but they can also lead to people not even trying to solo because they're afraid of being stuck in a lower-scoring draw if they fail. And for me personally, I think the fact that no one is ever discouraged by the scoring system from trying to solo makes the games more fun.

At first galnce though (and I may be misunderstanding), doesn't the lead based approach suffer from the same likelihood of an equivalent to draw whittling as in SC whittling? If someone is out in front with SCs, you may not want to agree a draw until you can team up with some others to drag them back a bit. Not sure it is the same, but it feels similar.


I have played both types of systems. It is definitely NOT the same.

In a draw-based system, a coalition of big powers could refuse to agree to a draw until they've beaten up enough on the smaller powers.

In a lead-based system, the only reasons someone would ever refuse a draw would be a table-topper wanting more centers, or someone who wants to form a coalition to beat up on a BIG power, not a small one. In either case, you would only refuse a draw if you thought you could get yourself closer to 18 or someone else farther away from 18. This is not the case in a draw-based system, where you could easily be willing to lose some centers if it meant someone else's elimination.
jay65536
 
Posts: 420
Joined: 10 Sep 2016, 18:13
Class: Ambassador
Standard rating: 1120
All-game rating: 1126
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby DQ » 23 Nov 2017, 17:14

super_dipsy wrote:Excellent summary of what is around and how they all work! Great job and very informative :)

I would be most interested in how you guage that and any stats you have. From many of the discussions here over the years, a common thrust has been that people want to play where more people go for a solo rather than the draw. Your name for the sort of system we have (draw-based) seems to imply that it will lead to more draws, whereas the comment about the lead-based system seems to suggest a bigger push for solos? (or is more aggressive a general comment rather than one on solos?).


My experience in Face to Face play is that people will head towards the incentives you give them. In 20 years at Dixiecon, where the best, reliable non-solo result was a three way draw, players were sizing up their partners in Spring 1901, and trying to figure out who on the other side of the line would be their third. Anything other than a 3WD was a disappointment, and absent a solo, you could expect to win the event with three 3WD - if you had the most dots among the various people who accomplished this. It made for games that were occasionally less than fun, especially for the people who didn't understand the meta. Carnage scoring is all about topping the board, and, absent a solo, you can expect to win if you manage that three times, as long as you have the most centers among the people who have done that, dot-count being a tie breaker.

My interest is because there almost seems to be a contradiction in what we see here on Playdip and what we might expect. We generally run at a higher overall percentage of solos than most other sites where I have been able to view the statistics, and also as judged against 'accepted' rates in postal versions of the hobby (seems like historically the average used to be a solo roughly 1 game in 3). Admittedly we do tweak the 'draws-based' approach by uplifting a solo considerably, so that a draw is worth a lot more than two 3-way draws, and it is also possible that because our system is Elo-like it can sometimes mean people know they have to solo to get anything on their ratings, but even so I am a little surprised our solo rate is as high as it is.


That is an astonishingly high ratio of solo victories, and from outside here, suggests a mix of people who know how to play the game and a sizeable pool of people who either don't, or don't care. The ratio in F2F is more like 1 solo in every 20 games. That is driven by a number of factors that IMO have to do with the environment. For example, at WDC in 2016, it became clear that the only real way to ensure you'd make it to the top table was to get a solo - and there were a disproportionate amount of solo wins in the qualifiers. One player who DID get a solo actually missed the top board, as he had no other result, and two players had higher accumulated scores (Good example of "how many good results are worth more than a solo" in action there). The game has changed a lot since the postal game - what's the rate of solos here? Are you saying it is higher than one in three? In 44 games of full press over the last two years, the group that plays over on Spark has I think five solo victories - ish, we haven't kept great records, and there is no scoring system in place. (That is changing in 2018!) I'm very curious as to what the players in the elite groups - the ones that limit membership to players who at the very least don't NMR / Abandon games - see as the solo-to-draw ratio.

But perhaps the question of an 'aggressive' game does not equate to a solo rate. Maybe people get too hung up on solo rates when they are looking for a 'good game'. Perhaps the lead-based approach would encourage more general aggression but not necessarily more solos (or actually, I suppose he who has the most SCs gets a 'solo' anyway?). Hmm - that is another question I have. I am not sure I understood right. In the lead-based system, if I get to 18 centres, do I get a bigger reward than agreeing a draw with 16 centres? But certainly an interesting idea, and a good discussion. At first glance though (and I may be misunderstanding), doesn't the lead based approach suffer from the same likelihood of an equivalent to draw whittling as in SC whittling? If someone is out in front with SCs, you may not want to agree a draw until you can team up with some others to drag them back a bit. Not sure it is the same, but it feels similar.


Not sure if this was answered below, but in most lead based systems, the solo gets ALL the points, and everyone else gets 0. One factor in F2F games is that if someone is being a bad actor by voting down the draw until someone else is cut out, they run a very sizable risk of having the entire table turn on them - I've seen larger powers write hold orders while the person who was insisting on a "better result" that moved them from say 5th place to 4th place was eliminated. I've also seen that result in a thrown solo - so its perhaps a riskier move where people can say "God's sake we've been playing for 6 hours, lets go get a pint instead of worrying about your score" than it is on the internet, where no one knows I'm a dog.
Stab you soon!
User avatar
DQ
 
Posts: 315
Joined: 11 Sep 2008, 14:29
Class: Diplomat
Standard rating: (1062)
All-game rating: (1063)
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby GhostEcho » 23 Nov 2017, 18:16

DQ wrote:...[W]hat's the rate of solos here? Are you saying it is higher than one in three? ... I'm very curious as to what the players in the elite groups - the ones that limit membership to players who at the very least don't NMR / Abandon games - see as the solo-to-draw ratio.


Earlier this Summer, for a project I was working on (that I never did get around to finishing/posting), I pulled results for what were then the last 300 standard rules games to finish on the site, which stretched from April 2016 through mid-June this year. I had initially hoped to get a big enough sample with no drop-outs, but after checking the first 20 results or so it was evident that probably a majority of games on PlayDip suffer from NMRs and surrenders.

A solo was actually the most common result (126, or 42%), with the next most common being a 3-way draw (95), and 2-way (34) and 4-way (40) draws making up most of the rest. The remaining 5 games featured 4 5-way draws and 1 6-way.

Since PlayDip allows draws without including all players, and scoring is based on size of draw (not centers), these numbers don't quite reflect the results that might happen in some other systems. All powers had a significant number of games where a non-soloing power survived but did not get a result, with ratios ranging from Italy (73 survivals without scoring, to only 59 draws) to Austria (44 survivals without scoring, with 70 draws).

Taking all 2100 powers played over those 300 games, 54% survived while 46% were eliminated. That 54% breaks down as 32% of the total scoring and 22% surviving without scoring. Of the 32%, 6% of total powers soloed and 26% drew.

Roughly speaking at PlayDip the "average" player could expect to score in a third of games, and get about four or five draws for every one solo.

One thing worth mentioning, I think, is game length. PlayDip doesn't allow games to conclude before 1905. In my sample, the most common game end was 1909; the median was 1910; half of all games ended between 1908 and 1912. My impression is that these are (significantly?) longer game times, in game years, than is common in the F2F hobby. Over my sample, games with a solo did not take noticeably less or more time than games ending in a draw.
Last edited by GhostEcho on 24 Nov 2017, 00:52, edited 1 time in total.
"When you absolutely don't know what to do any more, then it's time to panic." - Johann van der Wiel
"I'm not panicking, I'm watching you panic. It's more entertaining." - Elli Quinn
"[Diplomacy:] No dice or chance. Just calculated insincerity." - Counter Trap
User avatar
GhostEcho
 
Posts: 1840
Joined: 10 Aug 2008, 04:56
Location: Baltimore
Class: Ambassador
Standard rating: 995
All-game rating: 970
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby super_dipsy » 23 Nov 2017, 23:16

On the point of games with or without surrenders, I just ran off some figures to look only at games where there were no surrenders. In these games, 60% have ended solos. This is actually higher than the percentage for games where there were surrenders, but from a quick glance I see that some games that get ruined by surrenders end in draws because the games are so broken that the remaining players just want to get out of it, so they agree a draw for no other reason than to escape I guess.

Oh, I should have mentioned, I was only looking at ranked games.
User avatar
super_dipsy
Premium Member
 
Posts: 12068
Joined: 04 Nov 2009, 17:43
Class: Ambassador
Standard rating: 1000
All-game rating: 931
Timezone: GMT

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby DQ » 24 Nov 2017, 00:36

GhostEcho wrote:
DQ wrote:...[W]hat's the rate of solos here? Are you saying it is higher than one in three? ... I'm very curious as to what the players in the elite groups - the ones that limit membership to players who at the very least don't NMR / Abandon games - see as the solo-to-draw ratio.


Earlier this Summer, for a project I was working on (that I never did get around to finishing/posting), I pulled results for what were then the last 300 standard rules games to finish on the site, which stretched from April 2016 through mid-June this year. I had initially hoped to get a big enough sample with no drop-outs, but after checking the first 20 results or so it was evident that probably a majority of games on PlayDip suffer from NMRs and surrenders.

A solo was actually the most common result (126, or 42%), with the next most common being a 3-way draw (95), and 2-way (34) and 4-way (40) draws making up most of the rest. The remaining 5 games featured 4 5-way draws and 1 6-way.

Since PlayDip allows draws without including all players, and scoring is based on size of draw (not centers), these numbers don't quite reflect the results that might happen in some other systems. All powers had a significant number of games where a non-soloing power survived but did not get a result, with ratios ranging from Italy (73 survivals without scoring, to only 59 draws) to Austria (44 survivals without scoring, with 70 draws).

Taking all 2100 powers played over those 300 games, 54% survived while 46% were eliminated. That 54% breaks down as 32% of the total scoring and 22% surviving without scoring. Of the 32%, 6% of total powers soloed and 26% drew.

Roughly speaking at PlayDip the "average" player could expect to score in a third of games, and get about four or five draws for every one solo.

One thing worth mentioning, I think, is game length. PlayDip doesn't allow games to conclude before 1905. In my sample, the most common game end was 1909; the median was 1910; half of all games ended between 1908 and 1910. My impression is that these are (significantly?) longer game times, in game years, than is common in the F2F hobby. Over my sample, games with a solo did not take noticeably less or more time than games ending in a draw.


I'm not sure what to say - "my momma told me to not say nuthin if'n I couldn't say sumthin nice" - so, I'll just say that I understand why people want to form sub-communities where you require people to demonstrate that they won't abandon, surrender, or NMR.

As to length of game, my understanding is that most European f2f games play until 1907, 09, or 11 by convention. North American games vary more - most tournaments mandate that games cannot end before 1905. Where there is a central clock they tend to go to at LEAST 1908, often longer. Most NA tournaments don't have a mandated end time for rounds other than Sunday, so games have gone to 1920 and beyond when someone is trying to force a solo. Far and away those games are outliers. For timed rounds in NA, you tend to see a "range" of time between 5-7 hours normally - when the game will end, known only to the TD.

2/5 games ending in solo wins ... that's just a lot, right? How many of those wins came without an "asterisk" - an NMR/Abandonment? How does it work when a player abandons - Civil disorder? Do you find a replacement player? I see calls for those on the boards sometimes.
Stab you soon!
User avatar
DQ
 
Posts: 315
Joined: 11 Sep 2008, 14:29
Class: Diplomat
Standard rating: (1062)
All-game rating: (1063)
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby GhostEcho » 24 Nov 2017, 00:58

Noticed a typo. The middle half of games ended between 1908 and 1912. The full fourth quartile of games I charted went longer than 1912. Of those most wrapped up by 1915; all but three of the 300 were finished by 1920.
"When you absolutely don't know what to do any more, then it's time to panic." - Johann van der Wiel
"I'm not panicking, I'm watching you panic. It's more entertaining." - Elli Quinn
"[Diplomacy:] No dice or chance. Just calculated insincerity." - Counter Trap
User avatar
GhostEcho
 
Posts: 1840
Joined: 10 Aug 2008, 04:56
Location: Baltimore
Class: Ambassador
Standard rating: 995
All-game rating: 970
Timezone: GMT-5

Re: A (Choose: Great/Pointless/Annoying) Scoring System Thre

Postby jay65536 » 27 Nov 2017, 17:59

I will say this about my brief experience with the online draw-based approach. In FtF games, conserving time is much more of a factor than it is in an online game. In an online game, you can play a real endgame, which very rarely happens in FtF games because people have more incentive to end the game relatively early. Draw-based systems could potentially see more solos that come from botched draw-whittling than they could in an FtF environment, where you don't just have to ask "what is possible" but "what is possible in a reasonable amount of time". So I think there is a chance that using draw-based scoring in an online environment does not force the same "boring" style that you saw in FtF.

Basically, DQ, what I'm saying is, I remember very well the "pursuing a 3way from the beginning of the game" dynamic that Dixiecon often brought out. But what I'm saying is, are we sure that the draw-based scoring caused that? Or was it the draw-based scoring combined with the fact that long endgames are virtually impossible in a multi-round tournament crammed into one weekend?
jay65536
 
Posts: 420
Joined: 10 Sep 2016, 18:13
Class: Ambassador
Standard rating: 1120
All-game rating: 1126
Timezone: GMT-5

Next

Return to Diplomacy Lore

Who is online

Users browsing this forum: No registered users and 1 guest