Chingis
Joined: Jul 09, 2007
|
 Posted:
Feb 13, 2009 - 20:56 |
 
|
Yes, it's a numerical rating for each team, and there's a formula for it, but think a little more philosophically than that.
Is Team Strength:
a) An independent measure of the strength of a team in it's next match, and a way to ensure fair matches?
b) A part of the game, and a goal around which to plan one's team?
Which one it is depends on how it should be treated. If it's b), then we can pretty much carry on as normal, with the small proviso of stopping referring to the Blackbox as picking "fair" matches (but rather "even TS" matches).
If it's a), then why do we know the TS formula? Knowledge of how the Team Strength is assessed can only encourage the gaming of it. Re-jig the Team Strength formula to make it more representative, especially important now due to its use in the Blackbox. But don't let anyone know what the new formula is. Then everyone can get on with building their team as they wish, and the TS calculation (assuming it is fit for purpose) will help to ensure fair matchups.
There will always be flaws in the Team Strength formulation as a prediction of performance, simply because nothing's perfect. But if no-one knows what those flaws are, they cannot be gamed.
So, is it a) or b)? |
|
|
Peter_Thorpe
Joined: Oct 08, 2005
|
| Posted:
Feb 13, 2009 - 21:00 |
|
|
Snappy_Dresser
Joined: Feb 11, 2005
|
| Posted:
Feb 13, 2009 - 22:56 |
|
it is a poor calculation for A and is not in any way actually B
While I agree with you sentiment, I don't think the solution is to hide TS, as it will be impossible to do, and at best only make those with the ability to game TS a more exclusive group.
The solution is to go with a much more transparent TS, like TV, that is much less succeptable to silly exploits. |
_________________
<PurpleChest> the way it splooshed got me so excited
"I hear that shadow is a douchebag"
-Mr Foulscumm |
|
Hogshine
Joined: Apr 04, 2007
|
| Posted:
Feb 13, 2009 - 23:03 |
|
| Snappy's sig wrote: | | <TheSpoonyBard> snappy speaks truths |
I agree with Spoony |
|
|
Wreckage
Joined: Aug 15, 2004
|
| Posted:
Feb 13, 2009 - 23:07 |
|
| Chingis wrote: | . But don't let anyone know what the new formula is.
|
now there you say something what i felt what alot of people think on what i never agreed...
you know.... hiding the formular doesnt actually make people stop using it for their advantage... they just have to learn about the advantages by their expirience instead of looking at the formular... now will that gives game quality?
id say... as much as if you would hide the agility / strengh calculations in the game so yes.... if we seed games based on ts... its better we go the road a) then the road b) ..... anyhow.... teamvalue still is the best....since it doesnt matchups equal but balanced... taking all the other aspects besides winning into account aswell.... |
|
|
Frankenstein
Joined: Aug 02, 2003
|
| Posted:
Feb 13, 2009 - 23:11 |
|
| Snappy_Dresser wrote: | it is a poor calculation for A and is not in any way actually B
While I agree with you sentiment, I don't think the solution is to hide TS, as it will be impossible to do, and at best only make those with the ability to game TS a more exclusive group.
The solution is to go with a much more transparent TS, like TV, that is much less succeptable to silly exploits. |
I'm 100% with Snappy here. |
|
|
CircularLogic
Joined: Aug 22, 2003
|
| Posted:
Feb 14, 2009 - 00:17 |
|
| Snappy_Dresser wrote: | | The solution is to go with a much more transparent TS, like TV, that is much less succeptable to silly exploits. |
True... mus less succeptable to silly exploits. TV features official exploits, that are officially discussed and bring you inducement bonus. Yay! |
|
|
Eddy
Joined: Aug 04, 2004
|
| Posted:
Feb 14, 2009 - 02:51 |
|
| Chingis wrote: |
Is Team Strength:
a) An independent measure of the strength of a team in it's next match, and a way to ensure fair matches?
b) A part of the game, and a goal around which to plan one's team?
Which one it is depends on how it should be treated. |
The thing is... It's not "either/or".
The problem people have with TS optimization in [B] is that they don't realize that [B] is a different environment, with one more resource to manage. Since TS is used to match teams, it is logical that coaches do their best to ensure they don't "pay too much" relative to the actual efficiency of their team. The infamous +ST example stems from this: some coaches judged that the actual benefit of having +ST on some players is not enough when compared to its actual cost.
Of course, this leads to all kinds of abuse, which should definitely put TS into question. In theory, TS should be regularly revised to ensure that no skills/combos are over-/under-priced. Practically, this is neither doable nor wished by the powers that be (i might be mistaken on that, but it's the impression i've gotten).
So what does that mean? That TS is here and being abused, and this might partly account for the recent loss of interest in [B].
The good thing about TV, as Snappy suggests, is that, rather than judging the team's (supposed) STRENGTH, it judges the team's POTENTIAL. Every normal skill is worth the same, every doubles skill is worth the same, every stat increase (respectively) is worth the same. In theory, this means that if we both have the same TV, and my team happens to be much stronger than yours, then it's because i've made much better skill choices than you.
Does it stand the practical test, though? I'm not sure. I haven't played LRB5 (i'd love if people who have would chime in), but there are a number of glaring issues, though. First, the system values every normal/doubles/stat the same, but it doesn't care about the player on which you get it. In theory, if you take two rookie CD teams, they're supposed to be equal if one has a Block Bull and a +AG Hobgoblin and the other has a +AG Bull and a DP Hobgoblin. Maybe they are equal indeed, but i'm not sure. I know i'd pick the +AG Bull any time (but again, maybe i'm wrong. I'm sure there are much better examples available).
While some will probably think "well, it's BB, luck is part of it, just look at aging!", and rightly so, the problem comes from the fact that, if [B] uses TV, TV is considered to be 100% reliable. In [R], there's human control to ensure the numbers are not off. In [B], there wouldn't.
But back to the TV issues. Something i often read in the complaints about the TS in [B] is that it makes people turn down stat increases, typically +ST. As i've tried to explain, this is because they consider TS a currency to get skills, and that they manage their currency. The thing is: it's the same with TV. TV is also a currency with which you pay for your skills. All the skills/doubles/stats cost the same, but maybe the cost is not that well balanced. Of course, it is an official rule, but then, so is TR. And we all know that TR cannot reliably be used to have "fair" games. So maybe there would need to be a FUMBBL tweak to TV, but then, it wouldn't be official anymore.
What i'm trying to say here is that TV is probably not the solution to the problems (thought or existing), because, fundamentally, it works the same as TS. Granted, it would be much easier to calculate, and slightly less open to obscure abuse (which is why, personally, i wouldn't mind it replacing TS), but it would still be abusable. And you would also see threads about how taking +ST on a lino might be way too expensive in terms of TV, which was the basis of the last controversy. Actually, i can't be bothered to check right now, but i'm pretty sure such things already exist on TalkBloodBowl or the like.
To sum it up:
- you can't have a precise description of a team's strength for auto-matching without at the same time developing/encouraging a tendency to optimize the efficiency/cost ratio
- TV, like TS, is about paying for the skills you get. The better the skill, the more you pay.
- TV is official, but kind of rough as a descriptor, and maybe the prices are slightly off. TS is much more precise, but is also more open to abuse. There's a trade-off here, i think.
- One of the complaints was that TS "force" coaches into taking some skills/combos over some others (by making some too expensive and some others too cheap). I think the same thing is at work with TV, except that it's an "official" pressure. But, historically, i think it's safe to assume that FUMBBL has always proven independent enough to think outside of the box and question seemingly universal tactics. So maybe an official "tool of pressure" is not adapted to here.
As a conclusion, maybe it simply is impossible to automate team-matching for fair games, since i believe that, whenever you introduce a measurement system, the very system becomes another resource, to be managed like the others. Obviously, it's something that's disliked by a number of people here. I'd tend to agre with them (even though personally, i mostly ignore this resource).
In any case, i think the "picking" metagame is much less desirable than the TS metagame, so i'm ready to have the latter if it means i don't have to experience the former. Ideally, someone will come up with an idea with no drawbacks =) |
_________________
'The generation of random numbers is too important to be left to chance.'
Robert R. Coveyou |
|
Chingis
Joined: Jul 09, 2007
|
| Posted:
Feb 14, 2009 - 03:10 |
|
| Eddy wrote: | | As a conclusion, maybe it simply is impossible to automate team-matching for fair games, since i believe that, whenever you introduce a measurement system, the very system becomes another resource, to be managed like the others. Obviously, it's something that's disliked by a number of people here. I'd tend to agre with them (even though personally, i mostly ignore this resource). |
Precisely. Unless, if you don't want the measurement system to become a resource, you prevent the players from becoming aware of the measurement system.
Practically implementing this would be slightly more awkward than at present, but not terribly hard to do. People could suggest what what should be in a revised TS calculation in an open debate. But one trusted person (or small group of people) would take those ideas and create a formula. And not tell anyone exactly what it was.
Then, no-one knowing what the measuring stick was, no-one could use it as a resource. The only way this could happen is if players were able to reverse-engineer the new TS formula from the matchup suitability scores and the teamsheets of picked teams. If this was thought to be a possibility, simply stop displaying the matchup suitability scores!
A secret TS formula could be done, preventing both the "picking metagame" and the "TS metagame" if people were of the opinion that the TS metagame is an unfortunate, unintended consequence of trying to fairly rate teams.
I only raise this point as I see that there has been work to improve the TS formula, which is wonderful. However I'm aware that it can never be perfect, and that while hopefully a more accurate TS formulation may have less "gaming" potential, there will always be new ways to "game" a new TS. Only preventing people from seeing what these "loopholes" are will prevent deliberate and concerted "gaming" of them. |
|
|
Eddy
Joined: Aug 04, 2004
|
| Posted:
Feb 14, 2009 - 03:27 |
|
I don't think i agree with you here. Simply because i believe you've omitted a fairly obvious way of estimating TS: experience and match-ups.
Take the no-RR/Leader thing, for example. Maybe it originated when people looked at the TS formula, looking for loopholes. I'm not convinced it's the case, but for the sake of argument, let's imagine it did. The thing REALLY beame popular (to the point of infesting [B]) because people noticed it worked, and gave them easier games than if they played without it.
Not revealing the formula would most likely prevent people from spotting exploits beforehand, but powergamers would still figure out that when they took such and such skill instead of such and such, they were favoured. Especially considering that a new TS formula could not be very far from the current one (after all, Block + Dodge would still be more efficient than Diving Catch + Pass Block, and thus would logically be priced higher). |
_________________
'The generation of random numbers is too important to be left to chance.'
Robert R. Coveyou |
|
sk8bcn
Joined: Apr 13, 2004
|
| Posted:
Feb 14, 2009 - 14:05 |
|
IMO, TS works pretty ok. There's a few issues that would need to be resolved, but other than that, I find it works well.
I consider that, only seldom, a match within 10TS range is unfair (handicaps taken appart) |
_________________
Join NL Raises from the Ashes |
|
Eddy
Joined: Aug 04, 2004
|
| Posted:
Feb 14, 2009 - 14:15 |
|
If you've read the first post, you'll notice that the issue is not whether or not TS is "fair".
The issue is that some people (a fair share, i'd say) think that using TS means that a new resource is introduced in the game. This leads to balancing the TS-cost of skills with their benefits. The problem is that TS is fixed by the staff of the site, so, more or less directly, the staff encourages people to take such or such skills, and to play in such or such way. The typical, most publicly known example of this is the no RR + Leader thing. The rebate means that people are actually encouraged to only have the Leader RR, since they get easier games than if they had a more classical number of RR.
So, again, the question is more about the philosophy of using, or not, TS, and the things it induces, more than the fact that the formula as it is is accurate or not. A formula can be changed to be better balanced, but a flawed philosophy can't. |
_________________
'The generation of random numbers is too important to be left to chance.'
Robert R. Coveyou |
|
Purplegoo
Joined: Mar 23, 2006
|
| Posted:
Feb 14, 2009 - 14:18 |
|
Like many other things, TS does a better than average job until someone tries to break it.
In an environment like [R], you use TS and your better judgement to select a fair game. TS within 5-10 and your brain when it comes to mngs, handicaps or races or whatever works very nicely indeed.
When you're in an environment like [B], sad cases will try to break it. Which is boo. I'm not sure you'll ever get a formula that's unbeatable though, so TS is fine. IMO. |
|
|
Lithuran
Joined: Jun 01, 2007
|
| Posted:
Feb 14, 2009 - 14:25 |
|
TS is whatever you think it is.
There is what it was intended to be (A)
There is how some view it, both A and B or sombination of both.
There is how you view it.
There is no correct answer. TS is just a formula which was designed to give a rough measure of a team's strength. Anything more is speculation. |
|
|
Eddy
Joined: Aug 04, 2004
|
| Posted:
Feb 14, 2009 - 14:33 |
|
Of course it is speculation... If you prefer to see it this way, the question is: what should TS be?
And this question is essential to determining the future of B, for reasons obvious. |
_________________
'The generation of random numbers is too important to be left to chance.'
Robert R. Coveyou |
|
|
|
| |
Conceding v Goblins/...
Search
