8 coaches online • Server time: 05:36
* * * Did you know? The most valuable player is Thursdaynight Guitarclub with 96 MVPs.
Recent Forum Topics  Resolutions for BB? | Legacy Skills | Secret League Americ... |
2013-03-04 00:59:48
22 votes, rating 4.4
22 votes, rating 4.4
Piling On: Mathematical Analysis
I did some math to calculate how effective Mb, Claw, PO were. For those interested, I only looked at the probability of causing a casualty and assumed PO was being used anytime you didn't get a casualty on the first set of rolls.
It's easy to compute using the law of total probability. For example against 7 av with MB, we condition on the first roll)
P(cas) = P(cas|8+)*P(8+) + P(cas|7)*P(7) + P(cas|1-6)*P(1-6)
Note: the 3rd term is 0. The result is P(cas) = 14.4%
Here are the results:
It's easy to compute using the law of total probability. For example against 7 av with MB, we condition on the first roll)
P(cas) = P(cas|8+)*P(8+) + P(cas|7)*P(7) + P(cas|1-6)*P(1-6)
Note: the 3rd term is 0. The result is P(cas) = 14.4%
Here are the results:
| Av | No Mod | PO | Claw | Claw + PO | MB | MB+PO | Claw+MB | Claw+MB+PO |
|---|---|---|---|---|---|---|---|---|
| 7 | 7% | 13% | 7% | 13% | 14% | 27% | 14% | 27% |
| 8 | 5% | 9% | 7% | 13% | 10% | 19% | 14% | 27% |
| 9 | 3% | 5% | 7% | 13% | 6% | 13% | 14% | 27% |
Comments
Posted by lizvis on 2013-03-04 01:02:58
emo
Posted by neoliminal on 2013-03-04 01:06:55
I think it's already common knowledge that this is ridiculous.
Posted by Overhamsteren on 2013-03-04 01:43:09
Now do the same table with percentages to get a KO or better :O
Posted by Qaz on 2013-03-04 01:47:34
All in all life as a woodie has become much better. NO DP . 3% less going out chacne. and Journeymen!
Posted by pythrr on 2013-03-04 02:11:06
please work out.
Posted by seanh1986 on 2013-03-04 02:14:19
It is common knowledge that it was ridiculous, but I was curious mathematically, how ridiculous is it?
I also wanted to figure out for myself which skill-combinations are more dangerous.
I can work out the math for a KO, as well. I'll post that in another blog eventually.
I also wanted to figure out for myself which skill-combinations are more dangerous.
I can work out the math for a KO, as well. I'll post that in another blog eventually.
Posted by pythrr on 2013-03-04 02:20:34
thank you for reinventing the wheel
Posted by polardragon on 2013-03-04 02:20:56
I've seen the odds in various places before, but the Mb+PO vs Mb+Claw observation is thought provoking.
Posted by Topper on 2013-03-04 05:03:48
Done for you here mate :)
http://www.plasmoids.dk/bbowl/playbooks/On_Rampage.pdf
http://www.plasmoids.dk/bbowl/playbooks/On_Rampage.pdf
Posted by strikereternal on 2013-03-04 05:31:23
heh. you said 'anal'
Posted by Arktoris on 2013-03-04 05:55:49
well something has to make orc teams be afraid. And it isn't aging, fouling, or TR bloat from spp
Posted by Foad on 2013-03-04 10:15:00
I love the scientific conclusion of the mathematical analysis at the end...
"So, assuming that it is actually used, a 3-skill comb is more powerful than a 2-skill combo (except when it isn't)."
Wait... OMG... What about LRB4 Claw + RSC + Piling on!!!
38.45%
Someone call Buff Rowdy!!!
At the very least can we have an in-depth forum thread that I can troll please?
"So, assuming that it is actually used, a 3-skill comb is more powerful than a 2-skill combo (except when it isn't)."
Wait... OMG... What about LRB4 Claw + RSC + Piling on!!!
38.45%
Someone call Buff Rowdy!!!
At the very least can we have an in-depth forum thread that I can troll please?
Posted by jamesfarrell129 on 2013-03-04 10:51:09
Needs a graph.
Posted by Garion on 2013-03-04 12:04:59
chavo, that was 2 traits and you had to avoid ageing. Also it had a counter, so now where near as bad. The main problem with CPOMB is how easy it is to spam on all mens.
Plus I seem to remember a lot of people thought RSC Claw was broken too.
The change to claw is actually a great bit of game design though I think it should be -2 to a minimum of 7 av rather tha blanket 7. That way it would give poor treemen a little advantage.
PO is the main problem, allowing it re-roll and to stack was just stupid game design.
Plus I seem to remember a lot of people thought RSC Claw was broken too.
The change to claw is actually a great bit of game design though I think it should be -2 to a minimum of 7 av rather tha blanket 7. That way it would give poor treemen a little advantage.
PO is the main problem, allowing it re-roll and to stack was just stupid game design.
Posted by dode74 on 2013-03-04 14:22:45
"In particular, Claw+MB+PO causes a casualty in more then 1/4 of hits, against ANY av. That is ridiculous."
Should be 1/4 of *successful* hits. A 2d block has a 55.6% chance of success against an opponent with block, and a 30.6% chance against a blodger (unless you have tackle as well when it becomes 55.6% again).
Assuming your CPOMBer gets his maximum 16 blocks in a game, and they are all 2d vs non-blodgers (or he has both block and tackle as well as CPOMB), and he uses PO every time he doesn't cause a cas without it, and he manages not to get fouled off the pitch, that's an average of 2.4 cas per game.
Even if he manages to make every block against players without block or dodge (75% chance of a knockdown assuming he has block) it comes out at 3.24 cas per game.
Should be 1/4 of *successful* hits. A 2d block has a 55.6% chance of success against an opponent with block, and a 30.6% chance against a blodger (unless you have tackle as well when it becomes 55.6% again).
Assuming your CPOMBer gets his maximum 16 blocks in a game, and they are all 2d vs non-blodgers (or he has both block and tackle as well as CPOMB), and he uses PO every time he doesn't cause a cas without it, and he manages not to get fouled off the pitch, that's an average of 2.4 cas per game.
Even if he manages to make every block against players without block or dodge (75% chance of a knockdown assuming he has block) it comes out at 3.24 cas per game.
Posted by Reisender on 2013-03-04 15:57:39
now assume you have 4 pombers, frenzy and you get 3 dice blocks after half of the opposition is dead :D
Posted by dode74 on 2013-03-04 16:20:57
If you're able to use all 4 every turn, or even more than one a turn, then your opponent deserves to die ;)
Just to be clear we're talking about 4 players with block, tackle, claw, PO, MB and Frenzy? 4 legend beasts or 4 CDs all with doubles then?
Just to be clear we're talking about 4 players with block, tackle, claw, PO, MB and Frenzy? 4 legend beasts or 4 CDs all with doubles then?
Posted by happygrue on 2013-03-04 17:38:02
Come on dode, you are such a smart guy and yet you throw up these straw men so often. Rather than a silly assumption that you can pick apart, how about assuming that 30 such blocks a game with claw/mb/po and the right combination of tackle countering dodge or blocking dodgeless players. Any good higher TV chaos or nurgle player is going to get this in an overwhelming majority of games (or THEY deserve to lose). The expected value of that is... the other guy doesn't have a team left. What exactly are you arguing? That claw/mb/po doesn't kill people?
Posted by lawman on 2013-03-04 19:43:37
From what I have reasearched as well the top CDB blockers only average 1.6 cas per game. The only thing that they have in common is block/tackle/mb/po/claw. However, the highest cas/game blocker in the ranked division (1.18) actually doesn't even have claw. But has st4/pro/mb/po/frenzy.
Common other skills set seem to be Strength 4, pro, frenzy, jump up.
To me the most lethal would be block/tackle/ST4/mb/po/claw/frenzy/pro or jump up.
Common other skills set seem to be Strength 4, pro, frenzy, jump up.
To me the most lethal would be block/tackle/ST4/mb/po/claw/frenzy/pro or jump up.
Posted by pythrr on 2013-03-04 21:34:11
please work out
Posted by dode74 on 2013-03-04 21:40:39
"30 such blocks a game with claw/mb/po and the right combination of tackle countering dodge or blocking dodgeless players." - 0.55 * 0.27 * 30 = 4.5 cas per game, and I think your conditions are somewhat optimistic if you think you're going to get 2d blocks on the right people with the right people 30 times a game.
"What exactly are you arguing? That claw/mb/po doesn't kill people?" No, that it's not "ridiculous" as the OP stated.
"What exactly are you arguing? That claw/mb/po doesn't kill people?" No, that it's not "ridiculous" as the OP stated.
Posted by happygrue on 2013-03-04 22:25:06
I don't think they are optimistic at all. Of course you will have games where the elves dodge away and don't fail much. But you will also get pitch clears in fewer than 30 blocks sometimes.
4.5 cas per game is massive, considering that KOs are on top of that.
I'll point to Chuck vs. Blood bowl, which averages just shy of that in it's 300+ games. Now many of those are against poor teams or coaches, but that average is also including all the games in early development or (like at present) with a mostly beatup team and perhaps 0-2 clawpombers. It seems clear from an average like that that a good coach playing a team that is in good shape is going to expect to see more than 4.5 cas per game + KOs on top. It's hard to defend without a team left!
4.5 cas per game is massive, considering that KOs are on top of that.
I'll point to Chuck vs. Blood bowl, which averages just shy of that in it's 300+ games. Now many of those are against poor teams or coaches, but that average is also including all the games in early development or (like at present) with a mostly beatup team and perhaps 0-2 clawpombers. It seems clear from an average like that that a good coach playing a team that is in good shape is going to expect to see more than 4.5 cas per game + KOs on top. It's hard to defend without a team left!
Posted by dode74 on 2013-03-04 23:09:57
You point to one of the best teams played by one of the best coaches as an example of what's normal?
Perhaps it would be better to look at some actual statistics, such as those Koadah puts together ( http://www.cmanu.pwp.blueyonder.co.uk/bb/stats/stats.html ), which shows that yes, Chaos and Nurgle do cause more cas than any other team on average at high TV (there's a shock!), but at 1800TV+ they only average 1-1.5 cas per match more than other bash teams. Importantly, they don't win much more than other teams, with a win% in B at 1800+TV firmly in the Tier 1 bracket. Perhaps all those elves and skaven who are sitting above 55% in the box need a nerf ;)
"It's hard to defend without a team left!" - Seems that people are managing it perfectly well. Better, in fact, than they are managing it against the elven teams, skaven and even zons! More tackle required, perhaps?
Perhaps it would be better to look at some actual statistics, such as those Koadah puts together ( http://www.cmanu.pwp.blueyonder.co.uk/bb/stats/stats.html ), which shows that yes, Chaos and Nurgle do cause more cas than any other team on average at high TV (there's a shock!), but at 1800TV+ they only average 1-1.5 cas per match more than other bash teams. Importantly, they don't win much more than other teams, with a win% in B at 1800+TV firmly in the Tier 1 bracket. Perhaps all those elves and skaven who are sitting above 55% in the box need a nerf ;)
"It's hard to defend without a team left!" - Seems that people are managing it perfectly well. Better, in fact, than they are managing it against the elven teams, skaven and even zons! More tackle required, perhaps?
Posted by seanh1986 on 2013-03-05 05:36:07
my main point was that PO is the biggest contributing factor in making the casualty % go crazy high
Posted by happygrue on 2013-03-05 14:40:10
"You point to one of the best teams played by one of the best coaches as an example of what's normal? Perhaps it would be better to look at some actual statistics, such as those Koadah puts together"
Yes, by which Chuck Vs. Blood Bowl is a completely average chaos team in terms of casualties caused. The fact that the team has won so many more games than the average chaos team could be due to many different things. My theory is that most of the top coaches don't play high TV bash in the box. But if they did, they would have teams that looked a lot like Chuck.
But I guess we're at a point where we are going to disagree. You say expecting 4+ cas per game +KOs is not ridiculous, I say it is. That is just a difference of opinion about what it is to be ridiculous. I'm not saying that clawpomb is an awesome strategy that wins all games, I'm saying it's a ridiculous strategy that makes many, many coaches not even want to play the game. My comment about it being hard to defend without a team was not trying to imply that clawpomb is the best way to win - it means it's hard to PLAY THE GAME against these teams, because they are not playing the same game you are.
You say it's not ridiculous because it's not overpowered (from a win % point of view), I say it is ridiculous (from a design point of view) because it has caused so many coaches to throw up their hands and walk away from the game.
P.S. Please, for the love of Nuffle, will you get yourself a high TV box team and play with those guys before you argue this anymore? I know you are experienced, but I would respect your opinion much more if I saw you playing against the teams that you claim are not ridiculous.
Yes, by which Chuck Vs. Blood Bowl is a completely average chaos team in terms of casualties caused. The fact that the team has won so many more games than the average chaos team could be due to many different things. My theory is that most of the top coaches don't play high TV bash in the box. But if they did, they would have teams that looked a lot like Chuck.
But I guess we're at a point where we are going to disagree. You say expecting 4+ cas per game +KOs is not ridiculous, I say it is. That is just a difference of opinion about what it is to be ridiculous. I'm not saying that clawpomb is an awesome strategy that wins all games, I'm saying it's a ridiculous strategy that makes many, many coaches not even want to play the game. My comment about it being hard to defend without a team was not trying to imply that clawpomb is the best way to win - it means it's hard to PLAY THE GAME against these teams, because they are not playing the same game you are.
You say it's not ridiculous because it's not overpowered (from a win % point of view), I say it is ridiculous (from a design point of view) because it has caused so many coaches to throw up their hands and walk away from the game.
P.S. Please, for the love of Nuffle, will you get yourself a high TV box team and play with those guys before you argue this anymore? I know you are experienced, but I would respect your opinion much more if I saw you playing against the teams that you claim are not ridiculous.
Posted by dode74 on 2013-03-05 16:01:35
"Yes, by which Chuck Vs. Blood Bowl is a completely average chaos team in terms of casualties caused."
And not vastly better than any other bash team.
"My theory is that most of the top coaches don't play high TV bash in the box. But if they did, they would have teams that looked a lot like Chuck."
Maybe, maybe not. It's certainly a good build, but players like Anna Wu as a BC certainly help considerably. Everyone looked at Jill Roberts as the core of the team, but Jill builds are 10 a penny: Anna Wu was the difference, imo.
"You say expecting 4+ cas per game +KOs is not ridiculous, I say it is."
Actually I'm saying that only 1-1.5 cas more than your opponent per game is not ridiculous, particularly when it's not winning many more games than other races.
"You say it's not ridiculous because it's not overpowered (from a win % point of view), I say it is ridiculous (from a design point of view) because it has caused so many coaches to throw up their hands and walk away from the game."
Then maybe the game isn't for them. More players died under LRB 4 than under these rules - Koadah produced some stats which showed this: I'll try to find them.
"P.S. Please, for the love of Nuffle, will you get yourself a high TV box team and play with those guys before you argue this anymore?"
No, because I think Box (and TV matching) is itself flawed. When I do get the chance to play here I play Ranked as I think that's the closest to the "challenges" rules from LRB 6 we have access to. On Cyanide I play almost exclusively league games and have faced plenty of high TV killer teams.
And not vastly better than any other bash team.
"My theory is that most of the top coaches don't play high TV bash in the box. But if they did, they would have teams that looked a lot like Chuck."
Maybe, maybe not. It's certainly a good build, but players like Anna Wu as a BC certainly help considerably. Everyone looked at Jill Roberts as the core of the team, but Jill builds are 10 a penny: Anna Wu was the difference, imo.
"You say expecting 4+ cas per game +KOs is not ridiculous, I say it is."
Actually I'm saying that only 1-1.5 cas more than your opponent per game is not ridiculous, particularly when it's not winning many more games than other races.
"You say it's not ridiculous because it's not overpowered (from a win % point of view), I say it is ridiculous (from a design point of view) because it has caused so many coaches to throw up their hands and walk away from the game."
Then maybe the game isn't for them. More players died under LRB 4 than under these rules - Koadah produced some stats which showed this: I'll try to find them.
"P.S. Please, for the love of Nuffle, will you get yourself a high TV box team and play with those guys before you argue this anymore?"
No, because I think Box (and TV matching) is itself flawed. When I do get the chance to play here I play Ranked as I think that's the closest to the "challenges" rules from LRB 6 we have access to. On Cyanide I play almost exclusively league games and have faced plenty of high TV killer teams.
Posted by happygrue on 2013-03-05 18:51:52
Ah, when will I learn? :(