Hayati
Joined: Nov 07, 2006
|
 Posted:
Jan 19, 2008 - 20:23 |
 
|
Hehe yes- try to score in 3-4 turns not 2. It's very hard for Khemri to score in the remainder.
In the other half, just keep a square away, and tie up the Mummies as above. One hit on the ball carrier and you're done.
If you get a good kick off result, you can exert so much backfield pressure that they won't have time to retort. |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 19, 2008 - 20:31 |
|
| SillySod wrote: | | With context (i.e. in bloodbowl) I'd define it as the chance of something succeding or failing to achieve a particular in-game objective (maybe knocking a player over or scoring) and then it'd have a whole bunch more meaning based on the relative risks and rewards. |
But still, 'chance' is not defined specifically. Here is my go at it:
Then probabiltiy of an event succeeding is the number that the ratio of successes to total attempts tends to, as we approach infinite attempts.
Hence, aiming for a 6 on a single die, with n rolls, you would expect the number of 6s acheived to tend towards 1/6 as n gets larger and larger.
Now, back to BB- what does this tell me about my Khemri dodge? Nothing! It only becomes apparent after many many games, when the ratios level out. For a single event, it is meaningless. To date, nobody has developed any practical solution to the case of a single event. Hence casinos make money whereas most clients do not!
Example: if you complain about a double 1, remember that it is as equally likely as a 4 and a 5, or a 3 and a 2, or any other result. The problem is not the dice- it's the definition of success. And the general misunderstanding that probability applies to single events.
Finally, the stuff you mentioned- I did not even hear of an axiom until I was 18! I do not think these definitions are simple, for an average human being in this world. Well, not for me anyhow  |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
Russo
Joined: Apr 11, 2006
|
| Posted:
Jan 19, 2008 - 21:53 |
|
Well the Insurance Salesman seems to be taking a lead over the Math Student |
_________________
BAAA means NO!
[21:35] <@Purp|away> why is it whenever I come to check the PC, Russo is lowering the tone?! 
================================
[22:59] <princevaliant> NBL has gotten to be such a dirty channel
[22:59] <princevaliant> ty Russo
BiggieB is my idol |
|
Hogshine
Joined: Apr 04, 2007
|
| Posted:
Jan 19, 2008 - 22:06 |
|
| Prasomchai wrote: | | Example: if you complain about a double 1, remember that it is as equally likely as a 4 and a 5, or a 3 and a 2, or any other result. The problem is not the dice- it's the definition of success. And the general misunderstanding that probability applies to single events. |
Not quite, unless the order of those dice rolls matters.
But it is still true to say that while the probability is over an infinite number of events, it can still be used to work out which tactic would work best. For example, it is agreed that 2d blocks are better, if we take an example where no one has skills or rerolls, then you have a 3/4 chance of knocking him over, and a 1/9 chance of knocking yourself over. However, if he gets to choose the dice used (a -2d block) then there is only a 1/4 chance of knocking him over, but a 5/9 chance of knocking yourself over. However, this is only true for a very large (approaching infinite) number of trials. For the one trial you are interested in, you may as well just take the -2d block against that troll, not worrying about moving your players to help. I'm sure we all can see that this is not a good idea...
Casinos make money whereas <b>on average</b> clients do not because the odds are set so casinos win slightly. However, there is still a chance that any one specific client will make money. It's just very likely that he will make less money than others have lost before him.
And on axioms, I didn't hear of any until this year (19), but I would like to think that I had a pretty good understanding. Axioms just help formalise everything and communicate mathematical ideas far more easily.
Say, for example, that your blitz-ra was stood by a wood elf near their endzone, and the ball was nearby but in 2 tackle zones. So your blitz-ra dodges into the tackle zones, picks up the ball, dodges out and runs to the endzone. My probability tells me that this is unlikely (assuming no rerolls, there's a 1/108 chance I think), so I wouldn't do it, unless there was no other way to secure the ball, and moving that blitz-ra wouldn't allow the elves to score more easily.
I apologise for my writing style, I tend to just write as I think, so it ends up almost complete waffle. I also apologise for my maths if it's wrong in places. I'm hungry...
edit: I guess I should also add my own definition of chance, but I'm going to eat stuff first |
|
|
Britnoth
Joined: Aug 02, 2003
|
| Posted:
Jan 19, 2008 - 22:10 |
|
| Prasomchai wrote: | | I have a better win ratio than with any other race. I had a 33 game unbeaten streak, and lose only around 10% of all games. |
So, your tactic is to cherry sub 145 CR coaches? |
|
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 19, 2008 - 22:30 |
|
| Hogshine wrote: | | Prasomchai wrote: | | Example: if you complain about a double 1, remember that it is as equally likely as a 4 and a 5, or a 3 and a 2, or any other result. The problem is not the dice- it's the definition of success. And the general misunderstanding that probability applies to single events. |
Not quite, unless the order of those dice rolls matters. |
Yes I meant in the case where the order matters.
| Quote: |
But it is still true to say that while the probability is over an infinite number of events, it can still be used to work out which tactic would work best. For example, it is agreed that 2d blocks are better, if we take an example where no one has skills or rerolls, then you have a 3/4 chance of knocking him over, and a 1/9 chance of knocking yourself over. However, if he gets to choose the dice used (a -2d block) then there is only a 1/4 chance of knocking him over, but a 5/9 chance of knocking yourself over. However, this is only true for a very large (approaching infinite) number of trials. For the one trial you are interested in, you may as well just take the -2d block against that troll, not worrying about moving your players to help. I'm sure we all can see that this is not a good idea... |
Yes, I agree. However, it is good only in the long run. Therefore, when an AG6 player falls down on a dodge, you should not be surprised. Now, if you played 1,000,000 games and it happened every time, then that may be something!
My meaning was that you can use probability over lots of games, but in a particular game it will not really help you out. Ask the last Fling that killed a Troll
| Quote: |
Casinos make money whereas <b>on average</b> clients do not because the odds are set so casinos win slightly. However, there is still a chance that any one specific client will make money. It's just very likely that he will make less money than others have lost before him. |
Partly. But mainly due to the large number of risks that are pooled. Suppose that all casinos played only one bet, betting all of their estates, each with a 51% chance of success. Many of them would go insolvent. So a probable outcome itself is not sufficient. It must be coupled with risk pooling. The more pooling there is, the lower the volatility of the result. And therein lies the secret. |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 19, 2008 - 22:31 |
|
| Britnoth wrote: | | Prasomchai wrote: | | I have a better win ratio than with any other race. I had a 33 game unbeaten streak, and lose only around 10% of all games. |
So, your tactic is to cherry sub 145 CR coaches? |
LOL well assuming I did that with every race I used, then it still would stand that I have a better record with Khemri than with anyone else
[EDIT] Btw I have never refused a challenge, except when I already have a game. If the high CR coaches offer, I will play!
I think it can be a vicious cycle. In [R], many high CR coaches will not play a team with a good record, especially if the coach in question has a CR way below their own. So you end up playing low CR coaches, and improving the team record even more. |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!"
Last edited by Hayati on %b %19, %2008 - %22:%Jan; edited 1 time in total |
|
vanGorn
Joined: Feb 24, 2004
|
| Posted:
Jan 19, 2008 - 22:40 |
|
After checking all my teams' ratios, I have to admit that my khemri team is on top, just behind my chaos dwarf and righteous dwarf team. |
_________________
Gimme a pint of fungus beer!
Then we will climb the ladder.
 |
|
Hogshine
Joined: Apr 04, 2007
|
| Posted:
Jan 19, 2008 - 23:06 |
|
| Quote: | | Yes, I agree. However, it is good only in the long run. Therefore, when an AG6 player falls down on a dodge, you should not be surprised. Now, if you played 1,000,000 games and it happened every time, then that may be something! |
I think we'll agree on nearly everything! However, I do not agree that to coach Khemri successfully there is a greater need for a grasp of probability than to coach any other race successfully. There is a greater need to think carefully about positioning, but often this is done with no risk.
For example, my record with Khemri is better than my record with other races, but when I'm playing I do not think about probabilities. I know that the probability of me picking up the ball, or dodging away from someone, is 50% without reroll, 75% with. But beyond that, there is no reason for any serious probability. The probability I used here was GCSE (when I was ~16), certainly well taught by A level (~17/18 ). I am currently at university, and I can see no real use for the probability we are taught here being applied to Bloodbowl to help win a game. There is a lot about conditional probability (which would be helpful if I knew any future rolls, but irrelevant otherwise), distributions of probabilities (which is again useless) and the rest is basically proving things about probability using axioms and theorems. None of this is helpful.
Even in games where I do try hard, and plan everything I'm doing, I still don't tend to think of chances much. Sometimes I will think carefully about my chance to succeed in some move, and compare that to other options, but the only calculation I need to make is to multiply fractions. Khemri as a team only requires as good an understanding of probability as any other team, but it requires a greater amount of thought and understanding about risk.
edit: formatting to stop a  turning up |
|
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 19, 2008 - 23:14 |
|
Dude, feel free to edit the guide and replace 'probability' with 'risk' |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 19, 2008 - 23:22 |
|
| Hogshine wrote: | I am currently at university, and I can see no real use for the probability we are taught here being applied to Bloodbowl to help win a game. There is a lot about conditional probability (which would be helpful if I knew any future rolls, but irrelevant otherwise), distributions of probabilities (which is again useless) and the rest is basically proving things about probability using axioms and theorems. None of this is helpful.
|
Hehe, a nice summary of college there! |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
SillySod
Joined: Oct 10, 2006
|
| Posted:
Jan 20, 2008 - 14:57 |
|
| Prasomchai wrote: | | SillySod wrote: | | With context (i.e. in bloodbowl) I'd define it as the chance of something succeding or failing to achieve a particular in-game objective (maybe knocking a player over or scoring) and then it'd have a whole bunch more meaning based on the relative risks and rewards. |
But still, 'chance' is not defined specifically. |
I defined "chance" in the part above that. The with context bit was more an extra to show how I fitted that definition into bloodbowl. So people don't have to go back, here is "chance":
| SillySod wrote: | | Without context I'd desribe it as an event or collection of sample points (an outcome, such as the flip of a bias coin getting heads) which has a probability measure of 0.7 (chance of occuring is 7/10) given Kolmogorov's axioms (which basicly say that something must happen, the chances are between 1 and 0, and something about all the numbers have to add up if they dont cross). |
Now to take apart what you said for fun:
| Prasomchai wrote: | Here is my go at it:
Then probabiltiy of an event succeeding is the number that the ratio of successes to total attempts tends to, as we approach infinite attempts.
|
I'd say thats much more a statistical approach to a matter which is stricly probability since you only ever roll a given dice once. It is pretty much along the same lines as I dictated above though, just with tending to infinate attempts (which is safest to avoid, seriously, don't prod that shit with a barge pole) in place of a probability measure.
Ok, now the question, are axioms tough concepts? I'd say no. Each individual axiom is actually a pretty simple concept although they tend to be worded in a very precise way which can make it hard to see what that concept is. When you start looking at the implications and results of each axiom and how they interact it gets alot harder.
Now, for those of the unenlightened (/me makes a point of not looking at Russo, not even for a moment) what are Axioms? Basicly they're just some simple statements which are effectively "The laws of mathematics" which are the basepoint used to prove everything else. They can be changed freely (though not without consequence) to form whole different geomeries and mathematic systems but there are sets of "standard" ones which are used. 90% of them are just "obvious" statements like 1+1 = 2 (though that probably isnt one), the rest look complicated and are unobvious but mostly because you never even thought about them (secretly they're simple).
What has this to do with my point? Well, Russo may not know what an axiom is (he may do, I havent asked him) so dosent have a "strong grounding in probability" but is probably still capable of doing all the maths he needs for bloodbowl, if he could be bothered. It dosent really bug me but it is fun to argue about and it seemed like an obvious (if unimportant) mistake.
Time for some more helpful advice:
Take out the bit where you talk about your own winning record. Replace it by saying that Khemri can win alot whatever some people might think.
This guide will be read by two kinds of people: newbies and oldies (and possibly Britnoth). The newbies are looking for basic tips and will assume that anything written in the guide is golden, you don't need to persuade them otherwise, they dont want to read it and (most importantly) it won't help them to know that you have won x many games with Khemri. The oldies are looking for some interesting ideas or moves that they can adapt or try out (or just consider) to expand their gameplay. They don't want to hear how many wins you have because they don't care, worse, lots of us react in a similar way that Britnoth did and assume that you must have cherrypicked (its what I assume when I read that) which can actually undermine the percived validity of the rest of your article. Finally, if you write something which will be read by Britnoth don't give him free shots at it.
Keeping with the newbie, oldie (and Britnoth) approach, I'd recommend changing the style that you've written the article in. This will require a bit of work but if I was to write something for the guide I'd do it in one of three styles:
- re-writing the whole section so it either incorperated new points or was in a better format but taking care to credit the original authors (and probably move the original into another wiki and link it).
- adding an ammendment, briefly stating some observations and maybe a few extra formations or moves of note.
- adding another complete article
Each has its advantages and disadvantages. You've chosen the third approach, which isnt bad at all. However you might want to consider its target audience and how much you want to cater to newbies or oldies and therefore the style/format which is most appropriate as well as how much you want to rehash stuff which has already been covered.
Phew, long post  |
_________________
Putting the "eh?" back into Sexeh.
"There are those to whom knowledge is a shield. There are those to whom it is a weapon. Neither view is balanced." |
|
Hayati
Joined: Nov 07, 2006
|
| Posted:
Jan 20, 2008 - 15:11 |
|
Hehe, thanks for the comments dude!
I've removed the probability bit, as well as the reference to my own record, stating instead that many FUMBBL coaches have excellent records. It was there only to serve as evidence that the tactics worked is all- who was to know people would be so bothered
I added my bit as a new secton since it may well contradict the advice given previously. In this way, users can at least read all the sections, try every suggestion and make up their own minds. Also, I'd rather not poke and prod at other's work without permission!
Finally, Re probability, my point was merely that the definition I gave is the only one that you can relate practically to BB. And even then, it has severe limitations. But then we're discussing that damn topic, luck, and are away from the topic of this thread  |
_________________
"Hear me now!
All crimes should be treasured if they give thee
Pleasure somehow!" |
|
SillySod
Joined: Oct 10, 2006
|
| Posted:
Jan 20, 2008 - 15:25 |
|
| Prasomchai wrote: | | Finally, Re probability, my point was merely that the definition I gave is the only one that you can relate practically to BB. And even then, it has severe limitations. But then we're discussing that damn topic, luck, and are away from the topic of this thread |
I know, good isnt it?
Since you're presenting it as a seperate article you might want to put a comment right at the start saying that its organised into x sections with a very brief explanation of what each section is/contains. You might have already done this but I don't remember scrolling past it to read what you'd written. This is just so people know that its there and confirms that its seperate. |
_________________
Putting the "eh?" back into Sexeh.
"There are those to whom knowledge is a shield. There are those to whom it is a weapon. Neither view is balanced." |
|
Forces
Joined: Aug 05, 2007
|
| Posted:
Jan 20, 2008 - 15:43 |
|
Hmm...
The given chances just give you an idea. Of course even if the odds are against it every player could make a 6+ dodge, but still nobody would like to rely on them.
Even if an elve should fail every 6 dodge away from a dwarf no elf would stop dodging after the fifth.
So the chances of succeeding just gives you an idea what you should do an what not.
Exemple:
50% Chance of a skelly dodging and 83% chance for an elf, so this tells you it is better do dodge with an elf.
But still this does not need, that the skeleton fails every second dodge. On the long run propably but how much dodges will you do during one game?
But still you need to think a little bit more about what you do. If you missclick with a gutter who has dodge and ag4 and now needs to do a 3+ dodge it is not as bad as missclick and do a 5+ dodge with a blitzra.
And of course it is important to protect your ballcarrier. I watched a game were the ballcarrier were -2d in turn 8, resulting in two pows and still the mummi got the ball without reroll and scored... but as said above, it is not a good idea to rely on this.
Pick up the Ball and protect it at all cost (best do this by killing everything on the field) and hope that your enemy has no wiz^^ |
|
|
|
|
| |
Elf Draft Coach
Search
