### Clone Scoring

greenspun.com : LUSENET : MAME Action Replay : One Thread

I'm not sure if i am up to date with all the discusions on the new leader board scoring. I agree that advantages can be taken if there are multiple clones that play the same way. I like the percentage scoring a lot, much better than the 10 3 1. However i don't think doing away with clones and trying to determine what denotes a distinct clone, yes trog and trogp are different games but if you are good at one you will be similiarly as good at the other, enabling someone to get a double clone score.

I have a suggestion which probably isn't a solve all answer to the clone problem, but something to think about. Say You have game A with no clones and game B with 1 clone and game C with 2 clones, 6 possible games to play. Player A is good at game A and player B is good at game B, etc.

With the traditional Marp scoring. Player C would have an advantage of scoring three first places. Removing the clones would give all players an equal shot, however does not reflect the playability of all games available, since maybe player A likes very much playing a clone of game C more than the chosen parrent game.

What if we keep the clones but assign a percentage of the top score to each clone. So if C got first place on 2 games of C, and A got first place in her favorite clone of game C, then C would get a maximum percentage of 66%, and A could get a 33%.

The major problem of this is that it rewards stamina, which i think is a good thing, the ability to repeat a good score must count for something. But you do eliminate a major advantage for playing multiple clones.

-- Chad (churritz@cts.com), June 04, 1999

I have several ideas for the leaderboard...

Before I begin... I would like to suggest that all clones should be linked together into one game.(I think this has been suggested before... sorry for not giving appropriate credit.) For example, if person A scores 150,000 in Game A, and Person B scores 250,000 points in Game B, a clone of Game A, then Person B has a higher score than Person A. Wait a minute... what about games that score differently than others?(for example, a couple space invaders games, a couple amidar games) Well... you can either throw out those scores(which I think isn't the best idea...), or make it a separate game... or find a way to change the scoring so it's correct(which can be VERY difficult...)

Second, I think all ties should be scored as ties. In other words, either everyone gets first place points, or they get split together (I.E. if two people tie for first, they both either get 10 points, or the average of first and second, or 6 1/2 points(or to keep things whole... 7 points.).

Now I'll begin discussion of how the leaderboard should go.

1. Continue the original 10, 3, 1 system.

With 5,000 recordings coming up... I'm thinking of expanding leaderboards to five people, or maybe even more... here's some ideas...

2. Expand the original leaderboard, with a 100, 30, 10, 3, and 1 point system.

3. Because I think it's unrealistic... decrease the exponential scoring, to 25, 10, 5, 2, and 1 points.

4. Use a "Laff-A-Lympics" scoring system, expanded. 50 points for first, 35 for second, (25 for third, 15 for fourth, 10 for fifth), and 5 for sixth.(Those in parenthesses is what the original Laff-A- Lympics did...) This scoring system I would think would make things more competitive.

5. Give more points to those with more competitors. For example, the worst score gets 1 point, the 2nd worst gets 2 points, 3rd worst gets 3 points, etc., with the exception that second place gets one bonus point, and first gets two bonus points. This will give credit to all people involved.

Perhaps most effectively...

6. Use that percentage scoring system, discussed above.

7. Ditto #6 with a twist. Each game gets a percentage score, but they get added together, instead of averaged. For example, if a player gets 75 percentage points in Game A, 55 in Game B, and 100 in Game C, he would score 230 points, not 77(rounded), like MARP does... this would encourage those that play more games...

I don't know if any of these ideas will be effective... just giving my thoughts on this... thanks for reading this quite long post...

-- Gameboy9 (goldengameboy@geocities.com), June 04, 1999.

A better example of my origonal post, with the A,B,C players and games A,B1,B2,C1,C2,C3

A percentage addition scoring method. I think the margins will be screwed up but here goes.

Scores LeaderBoard Points Player A Game A = 50000 100 = 100*50000/50000 Player B Game A = 40000 80 = 100*40000/50000 Player A Game B1 = 10000 25 = (100*10000/20000)/2 Player B Game B1 = 20000 50 = (100*20000/20000)/2 Player B Game B2 = 18000 50 = (100*18000/18000)/2 Player C Game B2 = 9000 25 = (100*9000/18000)/2 Player C Game C1 = 300 33.3 = (100*300/300)/3 Player C Game C2 = 400 33.3 = (100*400/400)/3 Player A Game C3 = 200 33.3 = (100*200/200)/3 Player B Game C1 = 100 11.1 = (100*100/300)/3

The scoring would have to be floating point to be fair. Each clone would only be worth a portion to contribute to a whole. The only problem I see with this method is that you have to work extra hard to get the points for a cloned game where as if there happend to be only one base game you don't have to "work" as hard to get the full points. This sort of punishes someone for being good at a cloned game, forcing them to play more, which isn't that bad a thing is it? :)

I like the ties scoring method described by gameboy, but there would be a major problem with bowling scores, anyone who got a 300 would get full game points.

-- Chad (churritz@cts.com), June 04, 1999.