MattDakka
Joined: Oct 09, 2007
|
 Posted:
Jun 01, 2026 - 00:34 |
 
|
Assuming 2 Chess AIs playing (the same AI playing vs itself) the AI playing as White should win a bit more than 50% vs the AI playing as Black.
So, there is no luck in Chess, but a slight White advantage. |
Last edited by MattDakka on Jun 01, 2026; edited 2 times in total |
|
MercutioT
Joined: Mar 19, 2015
|
| Posted:
Jun 01, 2026 - 00:35 |
|
| RDaneel wrote: | In Blood Bowl, it should account for 20–25%... if we want to play seriously (majors... trophy)
If we want just to play fun ok . 50% |
What do these numbers even mean, though?
Really. Tell us. I will generously accept any reasonable methodology you propose, as long as you are consistent with it.
You're a very loud, passionate voice claiming that this edition is 'more dicey' and 'more random'.
What is your evidence of this? What are you seeing that brings you to this conclusion?
Fundamentally, Blood Bowl is extremely strategically oriented with respect to outcomes. There is a lot randomness, but much of it is best described as input randomness: Random things happen, good and bad, and most of the game is about figuring out how to best deal with or exploit those outcomes. For example, on offense, will I go south or north or will I not advance the ball at all? Well, I'll whiff on some blocks and succeed at others—totally randomly—and my offensive strategy is determined by what I have to work with. So how a given drive might play out is almost entirely unpredictable. But the final outcome—TD/victory—is incredibly well correlated to coach skill.
We all have our own quibbles and skepticisms of CR calculation. It's popular to say that 'CR is meaningless'. And we all know this or that coach who has an unhealthy investment in it. But the bottom line is that—as Christer has pointed out before at length—CR is an exceptionally good predictor of the outcome of a match. Obviously certain roster-matchups are weak against certain others, and that can undermine calculations. But in a mirror-match, at even TV, a 'Legend' coach is going to beat a 'Veteran' coach an overwhelming % of the time. I could comfortably bet $100 that it's 95% if not higher.*
And I would go further and say that even smaller differences in CR are still very predictive: Silly nomenclature notwithstanding, it is quite rare that your 'Megastar' coach is going to lose to your 'Star' coach, outside of certain tipsy team pairings. And nothing I've seen in BB2025 seems to have changed that—though it's certainly true that there are some new skills and tipsy team pairings that we might need to get used to.
So where do your numbers fit in here? What does your '20%-25% luck' mean? Do you mean that with no skill, you can still win a quarter of the time? Because that is certainly not true. If you were a child making moves at random, no amount of luck would carry you through. Even with the best dice in the world, you would lose 100% of the time. Which is notably not true in games that are actually highly random, like Incan Gold, Can't Stop, or individual hands of Blackjack.
*Happy to be proven wrong or right here by someone who's done the calculations. That money I bet can become 100USD towards a FUMBBL server |
|
|
Garcangel
Joined: May 04, 2006
|
| Posted:
Jun 01, 2026 - 01:21 |
|
| MercutioT wrote: | I could comfortably bet $100 that it's 95% if not higher.*
|
I think you would win that bet. That number is very close. |
|
|
koadah
Joined: Mar 30, 2005
|
| Posted:
Jun 01, 2026 - 13:08 |
|
| MercutioT wrote: | ...But in a mirror-match, at even TV, a 'Legend' coach is going to beat a 'Veteran' coach an overwhelming % of the time. I could comfortably bet $100 that it's 95% if not higher.*
... |
I don't know about RDaneel's numbers and not saying that these rules are more or less dicy.
But for context, back when I played in the competitive divisions, I figured that I might beat a top coach maybe 2 or 3 times out of 10 matches.
I don't see how much point there is in judging a game on matches played between Masters and rookies.
I am thinking that the 20-25% might be based on those people good enough that when they are playing a Master, they can "keep the game close and hope for a bit of luck". |
_________________
Official + Secret League rosters, old style skill progression, NO re-draft OR full 2016 rules. OR... 4000TV All Star Bowl - ALWAYS recruiting! |
|
MattDakka
Joined: Oct 09, 2007
|
| Posted:
Jun 01, 2026 - 13:28 |
|
Let's check the Box rankings (to remove the picking "statistical noise"):
https://fumbbl.com/FUMBBL.php?page=coaches&order=&type=blackbox&r=0
The top coach there has 75% win rate. I think that coach's skill must matter enough to have such a win rate.
It's not 80% or higher but it's high enough for a game with a great random element like BB. If only some strong races are played, it's possible to have 80% with them, by the way. For example, I think that it's possible to have 80% with Norse and High Elves. By playing only Goblins and Halflings, of course, that win rate is not achievable.
If more player's skill impact is desired, there are other games for that.
There are Dice Chess variants using two or three D6 to select the pieces to move or Coin Toss Chess, where a coin is tossed to decide which colour will make the next move. That's a mix of random and deterministic gameplay.
That said, I think that Blood Bowl got more dicey with the introduction of multi rr in BB 2020.
More dicey in the meaning you can roll more dice (often without caring about the best order of turn actions) and get less punished for doing so. With 1 rr per turn prioritizing the actions in the right order (or even avoiding to make secondary, less important actions at all) was more important than now. That doesn't mean there is no coach's skill involved now, on the other hand. There is still some coach's skill (enough to matter in the long run, odd games will still be lost by the best coach due to dice and/or match-up).
More or less, Blood Bowl always is dicey because it uses dice to work out the outcome of many actions and because of the automatic 1 failure, 6 success mechanic too, no matter how many modifiers there are. It would be hard to win more than 80%, even with a different ruleset, unless some mechanics were radically changed (for example, with more modifiers and less likely automatic failure/success).
I understand taking BB seriously, I take it seriously, just not too much seriously, because the rules are what they are. We can't expect too much from them. Let's try to do our best as coaches when we play and, if we lose, let's get not too mad about it. Being able to play Blood Bowl (as broken/badly designed a ruleset is) whenever we want without moving it's already a "win". |
|
|
RDaneel
Joined: Feb 24, 2023
|
| Posted:
Jun 01, 2026 - 20:22 |
|
| koadah wrote: | | RDaneel wrote: | | koadah wrote: | | Grod wrote: | | RDaneel wrote: |
I like to play a game where 80% of the score depend by the ability of the coach, not by chaos and randomicity. |
There is alway Chess. |
He does play chess and chess would be more than 80%, I think. |
no again
chess is HUNDRED PER CENT
... |
"No"?
Are you saying that you do not play chess or that 100% is not more than 80%
Edit: Or both? |
I m saying that in Chess outcome depends 100% on the coach's skill because there's no luck involved (you don't roll the dice), though of course there could be other factors, like your opponent eating a bunch of prunes and getting diarrhea—but let's not get into nerd's statistics.
In Blood Bowl, you roll the dice, but the increased chaos and incoerencies created by these 3rd season ruleset is increasing the RNG factor too much. Even for a dice game.
for sure this increase fun.
But it dampens my enthusiasm for competitive tournaments where you might have to study, train, and put in the effort, only to see it all go to waste because of ridiculous mechanics or absurd progression systems.
So let’s just play this game... as if we were at a casino... for fun. |
_________________
To judge a man, one must at least know the secret of his thoughts, his misfortunes, his emotions, Balzac |
|
JohnDaker
Joined: Aug 01, 2014
|
| Posted:
Jun 02, 2026 - 04:03 |
|
| RDaneel wrote: |
I m saying that in Chess outcome depends 100% on the coach's skill because there's no luck involved |
But that's not completely true. Whites have a greater win%. And luck involved into who is getting whites. |
|
|
MercutioT
Joined: Mar 19, 2015
|
| Posted:
Jun 02, 2026 - 06:28 |
|
| MattDakka wrote: |
That said, I think that Blood Bowl got more dicey with the introduction of multi rr in BB 2020.
More dicey in the meaning you can roll more dice (often without caring about the best order of turn actions) and get less punished for doing so. With 1 rr per turn prioritizing the actions in the right order (or even avoiding to make secondary, less important actions at all) was more important than now. That doesn't mean there is no coach's skill involved now, on the other hand. There is still some coach's skill (enough to matter in the long run, odd games will still be lost by the best coach due to dice and/or match-up). |
This is not how most people use the word 'dicey' to describe games. It's ok if you don't feel that missed rolls are being adequately punished, but that has nothing to do with an increase or decrease of randomness. In the context of games of chance, games with more dice rolling to determine the outcome are typically considered less dicey, because the decrease in importance of each individual roll. Blood Bowl typically involves about 400 dice per game. If a game (not Blood Bowl) used only 2 dice rolls to determine the outcome, we would typically describe that as 'very dicey', and if it used 1000 dice to determine those same factors, it would be less dicey—much less a game of chance, and with an inflated importance of whatever non-random elements the game might have.
Imagine two variants of Blood Bowl: one where both coaches have 0 rerolls for any reason, and another where they each have 30 rerolls and can use as many as they want per turn. The first is extremely dicey—just a couple bad pickups might decide the game. The second will be much more a game of skill; sure, a poor coach might succeed at some zany TD play by powering through with so many RRs. But their superior opponent can also do zany plays, along with having the knowledge of how to best use those resources effectively.
And luckily, we don't have to imagine these weird variants: BB7s has been a thing for years—a variant played with 7 players per team and (traditionally) no rerolls. It is famously dicey. And in the days of bloated CRP teams, there were plenty of 2000k squads with 6 or more rerolls. Which of these sounds more like a game of skill? If your life was on the line over a game of Blood Bowl, would you rather play BB7s or BB2020? I guarantee you would choose the latter.
Being able to use multiple RRs a turn is a definitive step away from randomness, and toward the paradigmatic Game of Skill, as described above. It's ok if you don't like it, but it's not less random. It's literally a reduction in randomness. At your discretion, your random bad roll can be ignored: reroll. It's right there in the name.
| RDaneel wrote: | In Blood Bowl, you roll the dice, but the increased chaos and incoerencies created by these 3rd season ruleset is increasing the RNG factor too much. Even for a dice game.
for sure this increase fun.
But it dampens my enthusiasm for competitive tournaments where you might have to study, train, and put in the effort, only to see it all go to waste because of ridiculous mechanics or absurd progression systems.
So let’s just play this game... as if we were at a casino... for fun. |
Ah, so despite my best effort, I guess we're not going to get an answer or explanation about what you mean by all your talk of 'increasing the RNG factor'.
I earnestly decided to do something that no one else on this site has done for you—Take your criticism seriously, unpack it, ask questions, and try to make sense of your perspective.
I won't make that mistake again. Best to take nothing said here seriously.
You're just ranting at the new thing.
I suppose it is the internet, after all.  |
|
|
CrisisChris
Joined: Dec 11, 2023
|
| Posted:
Jun 02, 2026 - 10:09 |
|
I can only assume that the reason for some coaches to complain is not based on 'Dicey' as you described it.
For me the overall assumption of the game was 'It is my turn, so I am in power to decide what to do and roll the dice'. Vice versa 'I set up a defence and make him roll dice'.
And I guess here comes the point: Multi RRs make defensive set ups useless sometimes. And that gives a bad feeling. It pretty much always is the case when you come to a conclusion that no matter what you do it will not have any impact.
For me that is the same with Steady Footing (and Foul appearance as well by the way). There is this bad feeling.
This does not mean that I share the oppinion, but I guess that is the root cause of the critics. (Correct me if I am wrong here) |
|
|
RDaneel
Joined: Feb 24, 2023
|
| Posted:
Jun 02, 2026 - 11:45 |
|
| MercutioT wrote: |
| RDaneel wrote: | In Blood Bowl, you roll the dice, but the increased chaos and incoerencies created by these 3rd season ruleset is increasing the RNG factor too much. Even for a dice game.
f |
Ah, so despite my best effort, I guess we're not going to get an answer or explanation about what you mean by all your talk of 'increasing the RNG factor'.
|
We’ve discussed the shortcomings of this third edition at great length. Go back and check out these forums. I’m not part of the majority, but I don’t think I’m far off the mark if I say that a good 30% of people consider this third edition (post-FAQ) a bit of a mess and RNG power increased.
You’ll find the reasons (not just mine) in below links
There are also other discords channels and stream where it was long discussed the limitations/criticism in bad mechanics and increase of RNG power of this edition.
This specific post forum was meant to discuss whether the May 2026 FAQ had at least fixed the main issues.
Well, no (in fact, just like with Grab/SideStep, they’ve introduced new )
https://fumbbl.com/index.php?name=PNphpBB2&file=viewtopic&t=33483
https://fumbbl.com/index.php?name=PNphpBB2&file=viewtopic&t=33548&highlight=
https://fumbbl.com/index.php?name=PNphpBB2&file=viewtopic&t=33526
https://fumbbl.com/index.php?name=PNphpBB2&file=viewtopic&t=33516
https://fumbbl.com/index.php?name=PNphpBB2&file=viewtopic&t=33528
Some people think it’s difficult to write balanced rules without errors and/or imbalances. Every role-playing/fantasy game is like that. I used to be a big fan of Magic: The Gathering; I remember the first edition from 1993 was full of errors that made the game basically unplayable—if you had certain cards and certain combos, you’d win in just a few turns. That’s why ban lists and subsequent re-releases were created. Today, Magic formats are such that if you play by the standard rules, the “luck” factor (the cards you draw) or the “dirty tricks” factor (creating deadly combos where you win after two turns) have been mitigated by a set of rules designed to ensure that the best players win, not the luckiest or the craftiest. For me, this is the ultimate goal of anyone writing a set of rules. And that’s why I think GW will have to step in during this third season to fix several issues that aren’t working if they want people like me to get excited about competitive tournaments again. |
_________________
To judge a man, one must at least know the secret of his thoughts, his misfortunes, his emotions, Balzac |
|
MattDakka
Joined: Oct 09, 2007
|
| Posted:
Jun 02, 2026 - 11:54 |
|
| MercutioT wrote: |
In the context of games of chance, games with more dice rolling to determine the outcome are typically considered less dicey, because the decrease in importance of each individual roll. |
More dice decrease the impact of each individual roll if you, for example, use 2D6 to work out a pass instead of 1D6. Given a finite number of rolls, 2D6 will be generally closer to the expected average than 1D6. The problem is that we roll many dice but many of them are single D6s. To make an example: you roll 4 rushes in your turn (forcing me to do the same to catch up your player). You don't roll 1s or manage to rr them. When it's my turn, I have to risk the same number of rushes and, since I'm rolling only four single D6s, I could snake the very first rush. The game shifted from "trying to minimize the number of rolls" to "roll rushes and hope". This makes positioning less important and, since positioning is related to coach's skill, I think that there is less coach's skill than before, with multi rr.
Warhammer used lots of dice rolls (way more than Blood Bowl, if you think that 2 archer units' volleys may be 20 dice rolls, then to wound rolls, then armour save rolls) yet was dicey, because not all the rolls had the same importance. While the average of a volley was generally ok, you could fail the Leadership test (based on just 2D6, not on 20D6). So, in other words, volume of dice roll doesn't necessarily equate to "less randomness".
Not all the dice rolls have the same importance. If I fail the very first rush (even with rr) and you don't I could not be able to stop your TD. Eventually, probably, my dice will average out, but it will be too late to matter.
Since with multi rr people are not punished as they should to roll dice mindlessly, the dice factor got more important over the positioning. I remember people more wary about risking single die actions with 1 rr per turn. Now it's not like that and there is less actions' order importance (which, again, is related to coach's skill).
| MercutioT wrote: |
Imagine two variants of Blood Bowl: one where both coaches have 0 rerolls for any reason, and another where they each have 30 rerolls and can use as many as they want per turn. The first is extremely dicey—just a couple bad pickups might decide the game. The second will be much more a game of skill; sure, a poor coach might succeed at some zany TD play by powering through with so many RRs. |
I think that the middle ground, i.e. limited rrs (say 3) with 1 rr per turn would be even more skilled.
With 30 rrs there is way less need to wisely manage the rrs' use, while with 0 rrs I agree that it's very easy to lose due to a variance's whim.
| MercutioT wrote: | | And in the days of bloated CRP teams, there were plenty of 2000k squads with 6 or more rerolls. Which of these sounds more like a game of skill? If your life was on the line over a game of Blood Bowl, would you rather play BB7s or BB2020? I guarantee you would choose the latter. |
I never said that BB played with 2000k squads and 6 or more rrs was a game of skill. I never played BB7, I don't like it for the simple reason it's more unbalanced than standard BB.
| MercutioT wrote: | | Being able to use multiple RRs a turn is a definitive step away from randomness, and toward the paradigmatic Game of Skill, as described above. |
I don't think so. My idea of "skill" is minimizing dice rolls, not rolling lots of dice hoping. Also, we should make a difference between rerolling 2D6 and 1D6. My main concern is about rerolling single D6s. That has, in my opinion, the highest gamebreaking potential because it's not that hard to snake, while a quad skull is way less likely.
| MercutioT wrote: | | At your discretion, your random bad roll can be ignored: reroll. It's right there in the name. |
Rerolling a failed rush doesn't guarantee I won't fail it again. I played lots of games with the scenario I described above: my opponent spamming rushes and I failing to do the same. Something that would happen less often with 1 rr per turn (because people would be more averse to rush spamming). |
Last edited by MattDakka on Jun 02, 2026; edited 1 time in total |
|
koadah
Joined: Mar 30, 2005
|
| Posted:
Jun 02, 2026 - 12:29 |
|
| koadah wrote: | | ThierryM wrote: | Looks like this debate is turning round between those old wise Sages trying to have fun with a pinch of Lore respected vs the hotheaded waac competitive driven coaches.
Hard to find a middle ground with goals so different between both parties. |
There are multiple discussions going on between multiple different people.
They are likely using the same words to mean different things.  |
Dicey: Mad, kick off table, weather, prayers, cas/removals etc.
Dicey: few skills, RRs. Bad dice mean ended turns & turnovers.
Dicey: loads of skills, RRs, +Stats. Actions likely to succeed. It was the 1s & skulls what won it.
Sweet spot. Goldilocks. Where it's "just right"
Everyone's "just right" is in a slightly different place. |
_________________
Official + Secret League rosters, old style skill progression, NO re-draft OR full 2016 rules. OR... 4000TV All Star Bowl - ALWAYS recruiting! |
|
Irgy
Joined: Feb 21, 2007
|
| Posted:
Jun 02, 2026 - 13:59 |
|
There is absolutely luck in chess it's just a different sort of luck. If there was no luck at all then a 1501 ELO player would beat a 1499 ELO player 100% of the time. You just have to be a little bit creative to work out what the luck actually is.
You make a move. It might be good it might be bad. You had reasons for making it, they might be sound or they might be flawed. You don't know how it's going to turn out. You might have lucked into the best move. It might have been a dangerous move but fortunately your opponent didn't find the response, but on another day they might have. The fact that you theoretically could know if it was good or not is irrelevant, because you're no more capable of doing that than you are of measuring all the air currents in the room to predict the deterministic physical action that is a dice roll.
The biggest difference really is it's just much easier to blame luck and whine about it in some games than others.
It's fairly hard to quantify "how much" luck there is in a game because at the end of the day if two people of equal skill play then the outcome is 100% luck. The degree of luck is entirely tied to the degree of variation in levels of skill. They're kind of two sides of the same coin. If the best player in a group wins 60% of the time is it because they're moderately more skilled at a game of skill or is it because they're extremely good at a game of luck? Is there really even a difference? |
|
|
MattDakka
Joined: Oct 09, 2007
|
| Posted:
Jun 02, 2026 - 14:20 |
|
| Irgy wrote: |
You make a move. It might be good it might be bad. You had reasons for making it, they might be sound or they might be flawed. You don't know how it's going to turn out. You might have lucked into the best move. It might have been a dangerous move but fortunately your opponent didn't find the response, but on another day they might have. The fact that you theoretically could know if it was good or not is irrelevant, because you're no more capable of doing that than you are of measuring all the air currents in the room to predict the deterministic physical action that is a dice roll. |
We should use Chess AIs playing as example, not weak human minds prone to mistakes. A mistake in Chess is not luck, it's a wrong or suboptimal move.
Modern Chess AIs are easily able to pick the optimal move turn after turn.
The right example is Chess AI vs Chess AI (the same AI).
In the end, it's a deterministic game and can be solved. It's just a matter of being able to calculate all the possible moves.
In Blood Bowl the optimal move (in terms of success %) can still fail due to dice, in Chess won't fail.
Given a board position in Chess, there is a cascade of optimal moves from there on.
A human mind can't calculate them all but an AI can (if not now, eventually it will be able to). It's just brute calculation of all the possible moves, picking the best branch of them.
Some moves are necessarily forced by other moves in Chess.
In Blood Bowl I can force an opponent to risk 3 dodges and they can make them all (especially with multi rr).
In Chess the King can't escape check by rolling Steady Footing or with Dodge. |
|
|
JackassRampant
Joined: Feb 26, 2011
|
| Posted:
Jun 02, 2026 - 17:46 |
|
The human mind acts as RNG in any game with human players, and any game without human players isn't functionally being played. Computers playing each other isn't a game, it's a simulation of a game. |
_________________
Lude enixe, obliviscatur timor. |
|
|
|
| |
The Stunty Cup
Search
