Amazing Video here, very relevant to Bloodbowl!

Everything dealing with the video game developed by Cyanide!
triperis
Posts: 26
Joined: 06 January 2017, 06:31

Re: Amazing Video here, very relevant to Bloodbowl!

Postby triperis » 01 June 2017, 09:46

Thanks all for more in depth discussions. I think that the key flaw of thinking "the system is good enough to achieve what we want" is that you kinda stopped on improving it. Given you just fiddled with coefficients ever since S1 shows that either you truly think there are no ways to game/improve it or stumble into confirmation bias, especially if you judge it by "minimum of complaints" (like people were complaining there was too much grind in S1, too few in S2 and rather good balance in subsequent ones).

If you count that at some point matching record X becomes easier from a fresh team than a developed one, it suddenly becomes advantageous to grind fresh teams until your winrate combines with luck "correctly". Then if we realize both luck and grind allow better positions come end of the season, how can we claim it is skill-based first and foremost? It would be if the better player would win 95+ % of games against a weaker opponent. But there is like no player, the best in BB2, who would have such records still playing people with slightly less than his skill. And with winrates <95 % strings of games we expect become very random.

It could be easily countered by allowing only 1 team by race. There are still many races to play if you wish, even on barebone version. And for trying to game your luck you get punished by being required to play races with which you are incrementally worse. But at the same time, it would introduce way more variety. This may be not a perfect solution, but this kind of considerations I find lacking, because ever since introducing the rating formula, you did not even try to look for exploits in it yourselves.

I do appreciate communicating and fiddling with the formula in general, even if people (me included) might critique your work, it shows your work is important for them ;)

User avatar
dode74
Posts: 7041
Joined: 11 December 2008, 11:18
Location: Nr. Reading, UK
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby dode74 » 01 June 2017, 10:38

how can we claim it is skill-based first and foremost
We're not. We're saying "team X had the best season under this system". "Best" is a matter of a number of factors, luck being one of them. Ability to play more games is another, as is skill.
It could be easily countered by allowing only 1 team by race.
I assume you mean per coach here? Anything which limits people's ability to play as much as they like won't happen.
Image

User avatar
VoodooMike
Posts: 1614
Joined: 14 July 2009, 07:44
Location: Calgary, Alberta
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby VoodooMike » 01 June 2017, 18:14

Given you just fiddled with coefficients ever since S1 shows that..
Going to stop you right there - while that formula was used for the first two or so seasons, it was changed completely after that. It's not simply knob turning and graph shaping anymore as far as I know - it is based on very specific breakpoints.
If you count that at some point matching record X becomes easier from a fresh team than a developed one, it suddenly becomes advantageous to grind fresh teams until your winrate combines with luck "correctly".
That's quite an assumption... where is that point, exactly? You're basing a lot of your argument on that idea so maybe you could take a moment to explain where it lies rather than hand-waving your way past these sweeping statements?
However I still argue that the current system of TV+ is a needless handicap on player skill as shown by games won.
It's not handicapping anything in CCL, it is matching you with teams that are closer in performance level.. since zSum (the performance aspect of TVPlus) and the ranking system are both based on wins and losses, you're not being artificially pushed down into the realm of lower-performance teams - those guys are also being matched with similarly performing teams when possible.

What you're really bitching about is that it's harder to artificially inflate your record against dramatically worse opponents in order to get fairly ridiculous 20-0-0 style records. The rankings are still showing us who the best teams/coaches are, but they are less subject to the sort of cheesing that many people relied on before.

If you're finding that under the current system you don't shine like you did under the previous system, it's simply an indicator that you were never particularly goood at playing the game, you were just good at gaming the game. That would be even more true under full TVPlus with the proper ranking system in place.
Friendly Reminder: Correlation does not equal Causation - tattoo it on the inside of your eyelids if it'll help.

Phage
Posts: 45
Joined: 08 September 2016, 01:46

Re: Amazing Video here, very relevant to Bloodbowl!

Postby Phage » 01 June 2017, 22:45


However I still argue that the current system of TV+ is a needless handicap on player skill as shown by games won.
It's not handicapping anything in CCL, it is matching you with teams that are closer in performance level.. since zSum (the performance aspect of TVPlus) and the ranking system are both based on wins and losses, you're not being artificially pushed down into the realm of lower-performance teams - those guys are also being matched with similarly performing teams when possible.

What you're really bitching about is that it's harder to artificially inflate your record against dramatically worse opponents in order to get fairly ridiculous 20-0-0 style records. The rankings are still showing us who the best teams/coaches are, but they are less subject to the sort of cheesing that many people relied on before.

If you're finding that under the current system you don't shine like you did under the previous system, it's simply an indicator that you were never particularly goood at playing the game, you were just good at gaming the game. That would be even more true under full TVPlus with the proper ranking system in place.
No that's not the issue at all. When someone with 5-0-0 record is searching for a match they don't get matched with someone of a similar record they get matched with someone of at least 250 TV higher then their team. This version of TV + in effect raises the floor of potential opponent TV regardless of win record. If you were matching on win record alone regardless of TV that would be different but the formula literally adds TV to the potential opponent pool.

User avatar
dode74
Posts: 7041
Joined: 11 December 2008, 11:18
Location: Nr. Reading, UK
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby dode74 » 02 June 2017, 05:15

No that's not the issue at all. When someone with 5-0-0 record is searching for a match they don't get matched with someone of a similar record they get matched with someone of at least 250 TV higher then their team. This version of TV + in effect raises the floor of potential opponent TV regardless of win record. If you were matching on win record alone regardless of TV that would be different but the formula literally adds TV to the potential opponent pool.
Because record differential has a TV "value" in terms of the odds of winning a match. It is, in fact, a more powerful indicator of who will win a match than TV is even with the current TV+ setup.

What you seem to want (and everyone else) is a limitless pool of people from whom you can be matched entirely on record. Well, we don't have that: we have the pool of people we have. That means that the best matchups in terms of giving both people as close to an even chance of winning as we reasonably can are already being made.
Image

User avatar
VoodooMike
Posts: 1614
Joined: 14 July 2009, 07:44
Location: Calgary, Alberta
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby VoodooMike » 02 June 2017, 10:32

No that's not the issue at all. When someone with 5-0-0 record is searching for a match they don't get matched with someone of a similar record they get matched with someone of at least 250 TV higher then their team.
The thing is it IS a similar record, you just don't see that. You think that it's important that they have the same number of games played and that the result is similar... well, that'd be ok too, but it turns out that the differential between wins and losses... not as a percentage, but as a simple subtraction, is a decent predictor of outcome... meaning that if we match on it, we reduce the pre-match bias.
This version of TV + in effect raises the floor of potential opponent TV regardless of win record. If you were matching on win record alone regardless of TV that would be different but the formula literally adds TV to the potential opponent pool.
Again, it covers that "win record", you just don't see it. Additionally, unlike win%, it's far less affected by a low number of games played, which is a major concern in a game where teams rarely last 10 matches.

Now, under CCL where your win% matters for ranking, you seem to think this causes problems... it also does not. What it does is try to match stronger teams with stronger teams and weaker teams with weaker teams. That doesn't make the weaker teams comparable to the stronger teams in rankings... instead, it helps the two types of teams better establish themselves in the rankings. When stronger teams face each other, they help us figure out which of them deserves a higher ranking than the other, which in turn improves the ranking's ability to determine which teams should go on to the finals.

What use is it to us if strong teams get matched with weak teams? I don't mean TV... TV is not a particularly good measure of a team's strength and certainly has no indication to the skill of the coach. A skilled coach vs. a weak coach is just the former padding his record... without it being about real accomplishment.... and THEN it becomes a random thing... lucking out with your opponents.

To a degree that still happens, but TVPlus rating matching tries to mitigate it. In no way is that "unfair" to skilled coaches, it just diminishes the ability of people to game the system. If we had full TVPlus it would basically eliminate it.
Friendly Reminder: Correlation does not equal Causation - tattoo it on the inside of your eyelids if it'll help.

User avatar
dode74
Posts: 7041
Joined: 11 December 2008, 11:18
Location: Nr. Reading, UK
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby dode74 » 02 June 2017, 13:13

Not my work:

Image
Image

Some more analysis by Schlice: http://www.nufflytics.com/post/the-value-of-tv/
Image

User avatar
VoodooMike
Posts: 1614
Joined: 14 July 2009, 07:44
Location: Calgary, Alberta
Contact:

Re: Amazing Video here, very relevant to Bloodbowl!

Postby VoodooMike » 02 June 2017, 20:30

Not my work
Good thing, too... since the guy whose work it is seems oblivious to concepts like third variable effects and variable interaction.

Is "win factor" synonymous with zSum? I'm not sure what those pretty graphs are meant to mean in terms of this conversation. Are those graphs of one CCL season or all the seasons combined? Did they look at each match from the perspective of BOTH teams, or just from the home team in each match? Hard to tell from that post, despite lots of pasted graphing code...
Friendly Reminder: Correlation does not equal Causation - tattoo it on the inside of your eyelids if it'll help.

licker34
Posts: 163
Joined: 09 March 2016, 17:40

Re: Amazing Video here, very relevant to Bloodbowl!

Postby licker34 » 02 June 2017, 21:30

I actually didn't think this vid was anything to do with TV+. Guess VoodooMike just has a bee in his bonnet for some reason.
No surprise there.
Cyanide/Focus have been tailoring BB2 towards the tournament scene with the Champions Cup and cash prizes, I thought that was interesting vs. dumbing it down like tabletop BB has with the BB2016 ruleset.
Have they though? They have added prizes, but what changes exactly have they made to 'tailor' anything towards the tournament scene?
There are kick off events that can decide the result of a match before it even starts, would it be a good idea to disable them for champions ladder?
Yes or no, it really doesn't matter, it's a ladder, the presumption is that everyone who wants to qualify will have to play enough matches to where the effects of these events should show up for everyone. Sure, there will be those who are hit more by them and those hit less, but it's not a one off match or even where losing/drawing one match means a team remake (as opposed to the WC system).

For the tournament itself there could be a stronger argument for reducing variance made.
A lot of competitive people would like less randomness but I'm sure a lot of casuals would like more.
I agree with the first part of that statement, but I don't know that it necessarily follows that casuals would like more. A blitz or a rock works both ways, it's not as though it happening to the casual when matched with a competitive gives them a warm fuzzy just because it existing 'works for them' more.

We already have upgrades which can remove portions of the KO table afterall. We also have presumed 1st half dicings which can be turned into interesting matches due to some wonky table result to start the 2nd half, but I wouldn't argue that as a reason to keep them myself.

As to the video itself it does raise some interesting points, however, I'm not sure how well it relates to a product like BB2. Firstly there's the aspect of 'skill' being differentiated between a 'reaction' game and a TBS game, and then there's the aspect of sponsors. Cyanide doesn't seem to have an entry to ESports, and (to my knowledge) no sponsors lining up to promote WC2 or whatever tournament scene they would like to promote. Even if they do, the size is tiny, so the monitization of the product is not based off of selling licenses to sponsors, but still to selling copies of the game to players.

As most of us have said at one point or another, the 'real' way to get your individual value out of BB2 is not by playing COL or even CCL, but by joining a league which caters to the specific way you want to enjoy the game (of course for some that way is COL or CCL). Thus cyanide may be better off focusing their efforts on improving league options and promoting the game that way.

Of course they can't really promote all the leagues, so they make their own (COL/CCL) and can promote through the CCL tournament, but that's still not really run by cyanide or focus. Their interest ultimately is to get as much publicity from the players and as much free work from the players as they can. Because at the end of the day, I think they still recognize blood bowl for what it is, a niche game that appeals to a minority of gamers. Recall that in BB1 they did include various other play modes, but those modes went no where. Perhaps they understood (or simply chose) to only allow 'traditional' play for BB2, even if that limits the appeal, trying to broaden the appeal apparently didn't work.

triperis
Posts: 26
Joined: 06 January 2017, 06:31

Re: Amazing Video here, very relevant to Bloodbowl!

Postby triperis » 06 June 2017, 10:12

It could be easily countered by allowing only 1 team by race.
I assume you mean per coach here? Anything which limits people's ability to play as much as they like won't happen.
It does not limit ability to play. You may play much more than 42 games, you may play any race. It limits minmaxing your luck for the record. If you think one should be entitled to play any race at TV1000, I say one should be entitled to play at any TV from the start. I would start at TV where luck factor would be the least influential (i.e. peak TV of the race), most would start at TV2000+ chaos. You will see the same mechanism at TV1000 when LE is introduced, where runs may as well be defined how lucky you were to avoid/dice amazons with a late-peaking team.


Return to “General”

Who is online

Users browsing this forum: No registered users and 9 guests

cron