dode74
Joined: Aug 14, 2009
|
 Posted:
Mar 17, 2016 - 00:37 |
 
|
licker - I've answered that question to the best of my ability already. I don't know what more you want from me, but claiming I "won't answer" it is false.
And yes, the golf example is perfect. There is quite literally no point in me playing you if I am going to get somewhere in the region of 110 and you are getting 210 - the game is a foregone conclusion. The entire point of the handicap is to prevent the game being a foregone conclusion so there is at least some aspect of competition about it. |
|
|
Matthueycamo
Joined: May 16, 2014
|
| Posted:
Mar 17, 2016 - 01:54 |
|
Basically TVplus is equalty of outcome and TV (provided you have a decent match maker and not BB2's) is equality of opertunity. |
_________________
DLE College 7s |
|
licker
Joined: Jul 10, 2009
|
| Posted:
Mar 17, 2016 - 03:03 |
|
| dode74 wrote: | licker - I've answered that question to the best of my ability already. I don't know what more you want from me, but claiming I "won't answer" it is false.
And yes, the golf example is perfect. There is quite literally no point in me playing you if I am going to get somewhere in the region of 110 and you are getting 210 - the game is a foregone conclusion. The entire point of the handicap is to prevent the game being a foregone conclusion so there is at least some aspect of competition about it. |
Is there really any aspect of competition though? Me claiming I beat you is utterly asinine in that context.
Have you answered though? You've said that you recommended TV++ to cyanide, but that's not my question.
Where is the data that supports that a majority of players want more changed than just large TV discrepancies? |
|
|
pythrr
Joined: Mar 07, 2006
|
| Posted:
Mar 17, 2016 - 03:40 |
|
| PainState wrote: | | pythrr wrote: | | i do not want to play tennis with my feet tied together just because my opponent is a retard who can't hit a forehand shot to save their own mother. |
What in the hell?
A glimpse into the life of pythrr, he plays tennis?
I always thought he was a golf type of guy. |
I am pan-talented. |
_________________
 |
|
dode74
Joined: Aug 14, 2009
|
| Posted:
Mar 17, 2016 - 07:51 |
|
matthueycamo - no, TV is some equality of opportunity; TVPlus is more equality of opportunity. Equality of outcome implies a result is imposed on you, and the result is not more imposed on you than any other odds-based result is imposed.
licker - yes, it's a competition. Far more so than if we know the outcome (to any degree of certainty) beforehand. That's truly not a competition.
And I've answered your question to the best of my knowledge. The assumuption within it, though, is that it's up to the players. If they don't like it they will vote with their feet. |
|
|
Strider84
Joined: Jun 03, 2009
|
| Posted:
Mar 17, 2016 - 09:16 |
|
| Quote: | | Me claiming I beat you is utterly asinine in that context. |
I think that is where you 2 talk about different things. Claiming i beat you would in your view imply you are better, which might no be the cas in tv++ (neglecting the luck factor for now) in a normal tv matchup whoever wins is better than the other guy (again given same dice)
So if you were to play for money, on a normal tv game you would bet 2 to 1 that you win if you think ur 70% favorite.
However if you play a tv++ match you would only bet 1 to 1.
So in general, a win in Tv++ is for the higher zsum player just worth more. As long as you think you can still increase your zscore you make money becaus youre better than the 50% edge. At some point people will know youre 70% and will not play you for money anymore except if they get 70 30 on their money.
As in a blind mm environment you cannot argue about the quota you are playing for, tv++ tries to even this out to 50 50.
A single game does not say who is the better player, but over some time the zsum should give a good estimate |
|
|
happygrue
Joined: Oct 15, 2010
|
| Posted:
Mar 17, 2016 - 13:47 |
|
| Strider84 wrote: |
A single game does not say who is the better player, but over some time the zsum should give a good estimate |
It may... but it may not work at all and perhaps even make matches less fair. What kind of a range will zSum have? What's the difference between players?
Does zSum go up to 30? Is a difference of 5 or 10 enough to measure skill? It seems not to me - as even a few-game streaks of luck or critical mistakes would be enough to skew this +5 for one coach and -5 for another. Suddenly they are matched up in a "fair" fight in which coaching skill is actually equal but one side gets 500K of inducements over the other. The result of that match then gets measured again in zSum... but did it actually improve anything?
Yes, if one coach has +20 zSum over another they are going to have shown themselves to be quite a bit better. But that's 1000 TV ! Do we really get to see such matches, or are we trying harder to stay in the range of 5-10 zSum? If the latter then a few points on such a scale will be relatively meaningless as they will bounce around over time.
This is my single biggest issue with this matching scheme - not the goal of 50-50 matches or metagaming or anything else: I just don't think it will actually work.
Moreover, I'll predict that it would be hated even if it DID work as planned.
Remember the Civ testers who kept complaining about luck when it was fair? Eventually the game designers shifted the luck, something like 20% to favor the players IIRC and then all the testers agreed that it was quite fair and balanced. When two khemri teams play the same roster but one gets 2 bribes and a wiz, even if it's "fair" from some point of view people are not going to go for it!
On the other hand, what do I know of the BB2 community. As pointed out by others, maybe this will work over there. I do agree that trying to make matchmaking more fair and reducing the "unfun" matches is a great goal so I hope this works out better than I expect it to. But I'm not going to hold my breath either. 
EDIT: Just to be crystal clear: my point is that random swings in luck will actually consistently produce unfair matches in such a system, which will then further throw off any ability to measure anything useful from zSum. Over time and many matches this could correct itself, but looking at the number of players, TV difference involved and average time a team lasts this is not going to happen for TV++, IMO |
_________________
Come join us in #metabox, the Discord channel for HLP, ARR, and E.L.F. in the box!
 |
|
WolfyDan
Joined: Aug 02, 2003
|
| Posted:
Mar 17, 2016 - 14:02 |
|
| Quote: | | And so it is said again, it does not matter if you end up at 50 50 after so or so many games, what matters if at what Zsum level you can maintain that 50 50 rate. |
But isn't your zsum dependent upon the result? Handicaps in golf work because your shot count is independent. Unless I am mistaken in the way that TV++ works (adjusted by win%), which I admit is entirely possible, then your zsum is dependent upon the factor it is controlling for. In that situation I would expect your zsum to be your model residual, which I expect will be TV synergy.
How are you controlling the zsum? Have you done any modelling to see how this would end up? |
|
|
dode74
Joined: Aug 14, 2009
|
| Posted:
Mar 17, 2016 - 14:29 |
|
| Quote: | | it may not work at all and perhaps even make matches less fair |
Not if you accept the 50-50 definition of fair. It can't. If the range is too large then it won't be too large by much, and will self-correct back towards the mean.
| Quote: | | Just to be crystal clear: my point is that random swings in luck will actually consistently produce unfair matches in such a system, which will then further throw off any ability to measure anything useful from zSum |
Randomness happens in every game, so the effect is the same. As with all such measuring systems the randomness is factored out over sample size.
wolfydan - TV was tested as a predictor as well, and was found to not be as good as TV++. TV isn't useless, though, which is why it is incorporate.
Cyanide have done some testing. No idea to what extent or what the testing involved. |
|
|
happygrue
Joined: Oct 15, 2010
|
| Posted:
Mar 17, 2016 - 14:48 |
|
| dode74 wrote: |
| Quote: | | Just to be crystal clear: my point is that random swings in luck will actually consistently produce unfair matches in such a system, which will then further throw off any ability to measure anything useful from zSum |
Randomness happens in every game, so the effect is the same. As with all such measuring systems the randomness is factored out over sample size.
|
I strongly disagree with this. A small change in zSum - 3 to be specific - is enough to give one team a wizard. My argument is that throwing on *another* metric, and one that is a lagging indicator of team performance on top of dice is going to produce (I expect, just guessing of course) very strange results that are never going to "balance out over time" because teams will get retired, players get maimed, rerolls are purchased, skills and stats alter the TV.
if a coach would normally win a game 60% of the time, but zSum is added to both teams and the zSum happens to be off even by a small amount such as 3-5 then you could be changing the chances of winning the game by a lot! Maybe that coach has a 70% chance of winning instead, or a 50%. Or worse!
If BB teams were anything like a stable, skill game such as the many that have been mentioned then this could have some chance of working. But since they are NOT I really suspect this will just add more noise into the system - and I have yet to see ANYTHING written to address that specific point.
EDIT: Of course, this assumes both players are good enough to take advantage of such an advantage. Maybe this actually works fine with a player base where the average player is not really very good and dice overwhelm the advantage of one team getting a bunch of "free" stuff. |
_________________
Come join us in #metabox, the Discord channel for HLP, ARR, and E.L.F. in the box!
|
|
licker
Joined: Jul 10, 2009
|
| Posted:
Mar 17, 2016 - 15:40 |
|
You know who knows more about the BB2 COL community?
Me.
Not Dode, not Mike not...
At least if the metric is actual games played in COL.
Now I have always said that participation is not a requirement for expressing an opinion. However, implementing a system in an environment you have no active participation in, and have not discussed with the community that will actually be affected by said system, is troubling to me.
And that's not a shot at Dode at all, it's a shot at cyanide.
Anyway, just to reiterate again...
Yes, TV++ works well when you look at the aggregate games played by a team. I strongly suspect it will be a disaster when you look at the individual games though.
And to get to the aggregate you have to play those individual games. What I and others are saying, is that within those individual games player skill gets washed out as the goal is to achieve the 50% win expectancy. Eventually the TV gap which TV++ 'ignores' means that the the game purely comes down to who got better dice. Skill no longer mitigates bad dice, nor abuses good dice. It's simply a grind to get though without failing too many rolls or watching your opponent succeed in their 'extra' rolls (or improved rolls).
You may have made it 'fair' by the TV++ definition of 'fair', but the point is that the assumption that 50/50 win rates was the 'fair' people were complaining about is incorrect. Or at least questionable.
We don't know though, since no one (meaning cyanide) seems to have bothered to ask the player base if this is what they want. |
|
|
Matthueycamo
Joined: May 16, 2014
|
| Posted:
Mar 17, 2016 - 15:44 |
|
| dode74 wrote: | | matthueycamo - no, TV is some equality of opportunity; TVPlus is more equality of opportunity. Equality of outcome implies a result is imposed on you, and the result is not more imposed on you than any other odds-based result is imposed. |
You think equality of opportunity is giving one coach more TV than another to play with?
It is giving two coaches the same mechanical tools and then they use their talent to get an outcome. Giving one coach more tools beause they have less talent is not equal opportunity it is a long term equal outcome with both hovering around the 50% win mark. That long term trend is imposed on a player regardless of talent. Best way of doing that is TV, not perfect because it never can be but it does the job well enough.
I get why you are trying to pretend it is though as it has more chance of selling that way than saying skill does not matter with winning matches long term anymore.
The problem with these matchups on BB2 is the vast TV diffrences that sometimes occour. This does not really solve that as a possibility and really could just make it worse with a larger TV team getting inducments. That is what most people seem to really hate when they complain about it when I read around. TVplus does not adress that, developing a new match maker could but TVplus with the same match maker is still going to have these wild TV differences people seem to hate so much. That is what I see as the fundamental problem. Even if it took longer to get a game the game would be so much better for it. |
_________________
DLE College 7s |
|
Strider84
Joined: Jun 03, 2009
|
| Posted:
Mar 17, 2016 - 15:51 |
|
I agree with happygrue. I have from the start mentioned that i think 50tv is way too much for the zsum factor. Lets look at it if it were 200, it would clearly lose any significance. Because noone could keep a zsum of 5 consitently. If it is only 30 though. You need to win 6 in a row to even give your opponent a wiz. So i expect a zsum of 10 to be possible to maintain. So given the luck factor of +-3 gamed i would say this can be significant enough.
Of course the point of median games of 6 comes into play, but then again, if you play your zsum 0 game, you might already get a wiz against a good tv110 team.
Or you might haveto play 100 flings which get a chef. But hey the're flings  |
|
|
dode74
Joined: Aug 14, 2009
|
| Posted:
Mar 17, 2016 - 16:23 |
|
happygrue - you're arguing around that same buoy again. If a wizard is enough to give a win then the zSum difference will drop!
FF, which some claim at a "larger amount" would do the same job, actually needs to be set to 60TV for it to do the same job. This isn't some "feeling" about what a wizard might or might not do in a game, it's a look at actual TV differences and an insight into what the effect is. Real data, not "but it's a wizard!" panic.
licker - "games played" is not the metric. You're making shit up as far as that goes, and cherrypicking to boot. Implementation of it might be "troubling" to you but, and I want to make this entirely clear: it's not your decision. It's Cyanide's.
matthueycamo - yes, I do, just in the same way as in golf you give the handicap. You and I are merely differing in our interpretations of what "equality of opportunity" means. If I were to play golf with Tiger Woods would I have the same opportunity to win, or is it a foregone conclusion?
The "long term trend" you are looking at is utterly irrelevant. You (and others) keep looking at win% as if it is the indicator of performance in this system and it simply is not. Skill does matter with winning matches long term, but it is not the win percentage you are looking to maintain, but the zSum. That's the paradigm shift I was talking about before, and I'm sure you or someone will say "yes, I get that, but the win%..." very soon, proving quite nicely that whoever says it does not get that at all.
You may have missed it, but there is also a pooling system coming. That means people will wait a little longer but get decent matchups, as you suggest. |
|
|
JohnnyFeyev
Joined: Apr 03, 2015
|
| Posted:
Mar 17, 2016 - 16:38 |
|
Can we please just leave the game alone if this is the only option? I'm so sick of where society in general is heading that we have to lower our standards to the lowest common denominator instead of raising them to achieve the greater standard. You only get good at Blood Bowl by getting beaten, sometimes destroyed. That's freaking life. The game already has an equalizer built in with the RNG. Why can't people see that the great coaches are the ones that can deal with getting diced and applaud them for it? Who are we trying to sell this horrible idea that everything should be fair to? The garbage players who will probably quit anyway when they get beat by someone who is better? Most of us that are active in the community have paid our dues. We got destroyed when we started, and we got better. Those that are passionate about the game sit around discussing the game, researching the game, experimenting with the game. We choose to become better. We don't want/need someone to tell us what is fair. Because fair seems to be favoring the garbage coaches. You will never learn how to get good at something until you lose consistently and learn from your mistakes. I am so sick of the everyone gets a participation ribbon, we don't keep score because we want everyone to feel like a winner mentality. There are good players and garbage players. There is nobody that thinks a garbage player should ever be able to beat a great player without the great player getting completely shanked by the dice. That would be like me being able to step into the ring with Mike Tyson thinking I have a reasonable shot at beating him. That's ridiculous! Just leave the game along, fix the matchmaking TV differential thing, and be done with it. |
|
|
|
|
| |
Blood Bowl 2024 Edit...
Search
