CorporateSlave3
Joined: Feb 07, 2004
|
 Posted:
Mar 01, 2005 - 21:01 |
 
|
And now, an observation:
I’ve been a member of one of the ever popular ‘MU’ groups for a while now, but please direct your comments about how evil they all are to another thread (there are plenty on the subject). I am here to discuss a different sort of ‘problem’ that I’ve encountered with these (and indeed other) types of groups – the scoring system.
Basically, the scoring systems used in many cases tend to favor (and perhaps ‘favor’ is not strong enough a word) the teams coached by people with a LOT of time to play Blood Bowl. The scoring system awards points for every game, which just keep getting added up. While there is something to be said for the award to go to the most ‘dedicated’ coach – i.e. the one who makes the effort and gets more games in – it hardly makes these often fun and ‘fluff’ filled groups worth the time for coaches who can’t play but for a few times a week.
The Living Rule Book actually notes this problem and structures a way around it (although their example applies to the suggested way to run a tournament): Simply have coaches only count their best 3 games and add the total. This may be too small a sample, and it might be best to count some of the worst matches as well…say add the top 5 and bottom 2, for example.
This would still favor coaches with more time – since they would have a bigger pool of games to select high scores from. However, it would give other coaches, who may be very talented but just short of time, a chance to make a respectable showing, rather than always get massively outscored by the team with time for 5 games a day.
It can be a bit disappointing to hunt matches for weeks, manage to get a few and do really well in the majority of them, then show up 250th in the points rankings. Adding some sort of scoring system that helps to mitigate the advantage that teams playing billions* of games have over coaches who have little time might actually increase the popularity of all sorts of points based groups on Fumbbl.
Any of the various Ranked groups are, after all, a very good way to get yourself and your team involved without having to adhere to the set schedules of a Tournament (which are also a challenge for the time-impaired coach). The more people that get involved and have fun on Fumbbl, the more fun it is for all of us!
Now, the floor is open to questions, comments, and flames of all sizes.
*Fine, so I exaggerate. Millions. |
|
|
Mully
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 21:11 |
|
| Quote: |
Basically, the scoring systems used in many cases tend to favor (and perhaps ‘favor’ is not strong enough a word) the teams coached by people with a LOT of time to play Blood Bowl. The scoring system awards points for every game, which just keep getting added up
|
Initial assumptionhere is false. For every game you play you can either gain points or lose points. SURE, someone who has played 100 games has the potential to have more points than someone with 10 games, but they also have the potentail to have less points.
I have played VERY few ranked games (compared to many) and have still cherry picked enough to stay in the top 150.
I agree the ranking system is broken (and a joke) but its not because of the formula used, it's due to the ability to cherry pick your way through an open environment. As long as "Ranked" and "Open" are linked together I think the issue will always exist. |
_________________
Owner of the REAL Larson
Come join the CCC League |
|
BadMrMojo
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 21:21 |
|
| Mully wrote: | | Initial assumption here is false. For every game you play you can either gain points or lose points... |
I don't think he was actually referring to {R} ranking but meta-league groups (like the *MUs) and their points systems, which do usually allow the possibility of a point loss but not a significant one.
That said, Mully has a good point. If you were to come up with a sliding scale which always averaged out to 0 points per match, then only the people who did well would end up with positive points and poor performances would result in negative points - no matter how many games were played.
That's a pain in the butt to maintain, however.
Another option is doing something like taking your average points in a set period of time (ie: each match, participants gain Match Points for their performance. Then, each week, all participants get League points equal to their average Match points/game during that week). Probably would work better with relatively frequent resets, I'd imagine. |
_________________
Ta-Ouch! of BloodBowl
Condensed Guide for Newbies |
|
Nichodemus
Joined: Feb 11, 2005
|
| Posted:
Mar 01, 2005 - 21:22 |
|
Great idea imo, as it will also allow a buffer for those of us who has a team that doesnt take the resuslts seriously, and only play each match for fun, as the results from these will be in the inbetween category (sometimes), and then wont pull as much down in the ranking  |
|
|
SnakeSanders
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 21:26 |
|
yeah i won 3 green horn matches and as a result my ranking went from 150 to 169!  |
|
|
CorporateSlave3
Joined: Feb 07, 2004
|
| Posted:
Mar 01, 2005 - 21:38 |
|
I think you've misunderstood, I'm not really saying anything about the Coach Ranking system, just the various 'points oriented' groups within the Ranked (or Unranked for that matter) divisions.
And while the idea that a lot of games gives the potential for losing more points, a good coach (or one who is good at cherry picking for that matter) who has lots of time will beat out a good coach who has little time, since they will invariably have many more points from having more 'good games.' Granted a n00b who plays a lot of games will probably do poorly so much of the time while they're learning the ropes that they will have fully expressed the greater lost points potential - but the system is not to reward coaches for poor play, but to reward well performing coaches who have little time to play.
While in theory more games = more chances to both lose and gain points, a 'good' coach will tend to gain more (and more often) than he loses. So given two very 'good' coaches, and given that one plays 30 games while the other plays 10, will have the coach playing 30 games ahead of the one who plays 10 I would say 99% of the time. Even if the shorter (gamewise) coach performs better, for example scoring 20 points per game vs. the longer coach scoring 10 points per game. The shorter coach did a better and more consistently better job, but got beaten out by brute numbers, which he never had a chance at in the first place.
In the mean time, it is true that a bad coach will end up with a consistently lower score, but that's how it is meant to go: that's how the teams with the 3 best games get into the semis in the LRB tournament system.
EDIT: Yeah, what BMM said (should have quoted in my post if I was gonna take so bloody long typing it)...not about [R]anking, but groups that keep score. |
|
|
monboesen
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 21:54 |
|
My first suggestion for the MU's would be that you get only positive points for games you won. Just to keep winning the number one priority.
And yes some sort of sample of peoples best and worst games would be a decent way of doing it. Another would be to take the average of all your MU games. But that would favorise the guy with one great game. Maybe an average that only takes into account games played within n weeks (with n being some agreed upon number). |
|
|
Arcon
Joined: Mar 01, 2004
|
| Posted:
Mar 01, 2005 - 21:59 |
|
The E.L.F. awards only points for a single match played against one race. This is a possibilty in giving the same chances to each one. I guess there are more ways to handle it. |
|
|
Mully
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 22:14 |
|
| Quote: |
don't think he was actually referring to {R} ranking but meta-league groups (like the *MUs) and their points systems, which do usually allow the possibility of a point loss but not a significant one.
|
Oh ......... I'll shut up now  |
_________________
Owner of the REAL Larson
Come join the CCC League |
|
SnakeSanders
Joined: Aug 02, 2003
|
| Posted:
Mar 01, 2005 - 23:21 |
|
Mully backing down and being embarrassed? Well i never?  |
|
|
zredna
Joined: Aug 02, 2003
|
| Posted:
Mar 02, 2005 - 03:21 |
|
I totally agree with monboesen!
| monboesen wrote: | | My first suggestion for the MU's would be that you get only positive points for games you won. Just to keep winning the number one priority. |
Its annoying to lose a team vs MUs when the focus on killing players instead of winning. But if a claw/rsc team would first consentarte on winning and second on killing U can try to fend them of by scoring.
Zredna |
|
|
Laviak
Joined: Jul 19, 2004
|
| Posted:
Mar 02, 2005 - 05:22 |
|
Probably a system similar to that used in the 'championship' ranking would make sense. Basically to get a high ranking there, you need to play a certain amount of games (10 or so i think), and after that, they're averaged. This stops people from playing just one game (say scoring 5 TDs, injuring 10 players) and having a pretty much unbeatable average.
Maybe 10 is too many for shorter tournaments, but the basic idea is a good one (in my opinion). |
|
|
monboesen
Joined: Aug 02, 2003
|
| Posted:
Mar 02, 2005 - 09:04 |
|
| Quote: |
Its annoying to lose a team vs MUs when the focus on killing players instead of winning. But if a claw/rsc team would first consentarte on winning and second on killing U can try to fend them of by scoring.
Zredna
|
The real problem is the easy excuses. "Bahh now your [MU target group] are up 2-0 I'm never going to win. Guess I can just as well kill and maim as many of them as possible" or "Who me, Yeah I play to win. I just suck at the game. Its not my fault".
I've specced enough of the various MU coaches games to believe that at least some of them act like this. Requiring a win to get anything but negative points would go a long way to prevent that. It won't stop the senseless maiming but it may dampen it a bit and reinforces that the point of the game is to win by scoring at least one TD more that your opponent. |
|
|
sk8bcn
Joined: Apr 13, 2004
|
| Posted:
Mar 02, 2005 - 09:45 |
|
forget about *MU beeing right or not.
Arcon you are wrong since *MU is about to hit ONE race.
Corporate, I can't agree with the way you are proposing the scoring system. Isn't that going to make the whole thing die somehow? I mean, if 3 games are enough to get the maximum of points then why should I play more? Isn't that going against the spirit of the *MU's? (just to let everybody know: I do not belong to any */MU nor Kaos)
Besides, the bad behavior of killing teams would get higher. You will need to get x cas to get over team y, so you would have to destroy the opposing team. I couldn't work properly IMO.
I do really think that there is no chance for the points system to be improved else something alike 1 game recorded per week but that do not really match with the free system of such leagues.
I think you should accept it and, well, do your best with the time you have. |
|
|
paulhicks
Joined: Jul 19, 2004
|
| Posted:
Mar 30, 2005 - 01:46 |
|
i have to agree with corp here.... the mu's do seem to suffer this one problem with the scoring that it comes down to "ive got the most time, dp and claw/rsc so i win". i can honestly say this as it has alowed me to do so well in emu before i left. it was certainly a problem with dmu so i proposed the following changes after a conversation with corp.
all matches to be reported and the average score is the dmu total (eg 3 matches 10pts 15pts and 2 pts would =9pts) this would make number of matches less important than quality
new score system would be similar to last but with 2 sets of scores. first for normal dwarf hunters then an adjusted 1 for elven + stunty teams to try to even things up
and finaly points to be made even for doing well/ getting bashed. points doubled for a win, halved for a loss or draw.
of course dmu is over so i wont get the chance to try it but i thought it might have helped. |
_________________
Spelling, grammer and sense are for noobs! |
|
|
|
Roster Tiers
Search
