VoodooMike
Joined: Nov 07, 2010
|
 Posted:
Dec 15, 2011 - 19:58 |
 
|
| PainState wrote: | | In the end I feel that the math failed in this specific instance...what was needed was the old fashioned eye ball test and worry about what was happening in the "field" and not on the math calculators. |
This falls under the category of perception bias. You don't notice all the good drivers, or even average drivers, because you're completely absorbed in your own experience - you notice the bad drivers because they interfere with your experience in a way you notice. If you then conclude that there are a ton of terrible drivers, you're probably wrong, even if you've used what you think is unbiased intuition to determine that.
CLPOMB appears to be the same thing. People notice it because it negatively impacts their selfish personal experience in a way that they don't necessarily forget in 10 minutes, like most things. That doesn't mean its more prevalent than the math says it is, or that it has an objective effect beyond what the math says it does. It means that it has a high subjective effect.
So no, the math is not wrong at all. CLPOMB does not have the profound effect that many people feel that it does. So the real question becomes this: does it have a big enough subjective effect on the enjoyment of the game for us to give a crap about it, in spite of the fact that it does not have a significant mechanical effect on the game as a whole.
| PainState wrote: | | WOWZA!!!! and they have free govt internet hook up in their free apartment, playing on a free Blood Bowl sight. |
Not even close to being outside the realm of possibility, really. I run a chat site, and have seen plenty of heavy, proficient internet users who are, in fact, unemployable and living completely on government assistance. That said, most of them are still able to identify various categories of irony. The problem in Kelkka's case is, as we both know, one of deliberate obtuseness in the pursuit of conflict. |
|
|
f_alk
Joined: Sep 30, 2005
|
| Posted:
Dec 15, 2011 - 20:12 |
|
| VoodooMike wrote: | | The problem in Kelkka's case is, as we both know, one of deliberate obtuseness in the pursuit of conflict. |
| Kelkka wrote: | How did you come up with that number? Gut feeling maybe? Or perhaps eyeballing?  |
I would advise to look up the meaning of the symbol "" |
|
|
VoodooMike
Joined: Nov 07, 2010
|
| Posted:
Dec 15, 2011 - 20:35 |
|
| f_alk wrote: | | I would advise to look up the meaning of the symbol "Wink" |
Or I could contextualize based on more than one line from one posting. I could probably contextualize what it means when someone who has previously been snotty to me shows up in a thread with nothing to contribute but additional snottiness. Would it make you feel better if I added a winky icon to that? |
|
|
Kelkka
Joined: Aug 14, 2009
|
| Posted:
Dec 15, 2011 - 23:42 |
|
I just love it how defensive you are about your own use of vague numbers and call it "generalizing" and whatnot, and lay smack to anybody else who does the same.
I thought the whole point was how statistics were sacred and you can't just pull figures out of your pocket.
Well anyways, I go back to my gov paid rehab house for obtuse douchebags who try to learn how to tie shoe laces 
And the reason I didn't try to say anything constructive was that in all 3 recent threads (2 started by you and one hijacked) you have had plenty of debate, comments, ideas and feedback. Yet in every single time you have took a shot at pretty much everyone, you have not shown slightest attempt to actually listen or accept others' ideas or thoughts, and you pretty much just want to arise argument for arguments sake. I don't feel I could've said anything more constructive than pointing out double standards you have. And from your reaction I get the feeling the point wasn't too far from reality. Have fun arguing |
|
|
licker
Joined: Jul 10, 2009
|
| Posted:
Dec 16, 2011 - 00:17 |
|
| Kelkka wrote: | | you have not shown slightest attempt to actually listen or accept others' ideas or thoughts, |
For the fun of it...
Why should he accept others ideas when they don't relate to the topic he is interested in discussing?
Or...
If they are simply incorrect.
VM is an alpha anyway, most everyone else may be a delta bravo, but hey, that's the pecking order. |
|
|
Nelphine
Joined: Apr 01, 2011
|
| Posted:
Dec 23, 2011 - 00:15 |
|
I'd like to throw in my two cents; I haven't been around the site for several months (although every few weeks I poke my head back in), because:
I was a ClawPOMB victim. It caused me lots of grief in my 100-150 games played, and I just didn't like it any more. So I left. But I still love Blood Bowl.
This thread has made me want to post again, specifically for VoodooMike:
You have posted the tiers of various teams, in blackbox, and you have based that on the games that are played, and the winning percentages over those games played.
For a player like me, who did not have that many games played (although some might argue a 3 digit number of games is significant), not all of those games mattered.
Specifically, games in the 1000-1200 TV range did not matter. If I lost a game here or there, I didn't care; those games were where I could build my team up to a TV I was interested in playing.
The games from 1200-1500 TV are the games that mattered, and they are the reason that I left. In this range, I found that I faced numerous clawpomb teams (in an absolute sense; numerous in this case is defined as say, more than 5 (I think the number that did me was closer to 20, but it might not have been that high). If I played 5 games, and they were all clawpomb, then that would be numerous. Similarly, if I played 100 games, and only 5 were clawpomb, that would still be numerous.)
If I was playing a team where I was trying to get a few skills on the team, then I was bound to lose those skilled players to clawpombers, because that is what is advantageous for my opponent at the time. This resulted in me having some players that were bloat, and generally losing all my rare players (+S, blodge, etc), within a few games of gaining them. Partially this is due to poor coaching ability, especially in positioning, but if I had not played against numerous clawpombers, then this wouldn't have happened.
Then I considered going back to my original team (lizardmen with an agi 5 passing skink in Ranked); and due to the bloat on that team (I didn't know better when I was playing them), I was at a TV range almost entirely inhabited by clawpombers (2000-2300 TV), and I was literally afraid to play my team, due to my experience in the box.
Why does any of this matter to VoodooMikes information?
Because at 1000-1200 TV, if a team wins or loses, they are just as well off as they were before the game, regardless of what happens during the game. Therefore in this TV range, winning is the only thing that matters.
But if you like to play with a more diverse set of skills available to you, you need to have skilled players; for this you need to play at 1200 TV+ (and really, 1600 TV+ due to the necessity of getting block early).
So if you play a team for pure winning, you can look at nothing but overall win percentage, as VoodooMike has done. But if you want to play a team that will consistently stay within a certain TV Range that is not 1000-1200, then you need to look at 2 things: First, will the team do well in that range? And second, will the team stay in that range?
For this reason, I suggest that while VoodooMikes data is entirely valid for his definitions, his definitions are actually the problem. We should be examining 2 different ranges:
All games played (which will include 100's of games played by first time coaches, which we will have no way of accounting for, as well as teams that are going to be retired immediately, or played on a dare, or for whatever reason will possibly not be played with the same expectations as most games. of course 100's of games will not change the data much; but if it is actually 1000's of games, it could be significant, and we have no way of knowing whether these games are included.)
All games played above 1200 (maybe 1300?) TV - this would indicate that the coach was serious about playing with the team as a long term project, and would (I think, and I don't have data of my own to back this up) discount the possible problem games I mentioned above. It would also come closer to showing whether or not clawpomb (as opposed to teams who have access to clawpomb) has an effect on winning, as clawpomb cannot generally be found on teams under 1200 TV, and the Black Box is generally where we expect to find the most clawpomb.
If clawpomb does have an effect on winning, then we would expect to see high AV teams without clawpomb falling past 1200 (or 1300) TV; and teams with clawpomb increasing their win ratio past 1200 (or 1300) TV.
We would also be able to see how many games this was (out of the 108000 I think someone said VoodooMike was basing his data on); if it were high enough (say 30000? 10000?) then we could say that yes clawpomb does have a significant effect, of some kind. If there were extremely low numbers (less than 10000? less than 2000?) of games above 1200 (or 1300) TV, then we could say that regardless of clawpomb, even in an environment like black box where people can expand their teams as much as they want, their simply aren't enough such games to matter. But I believe there have been enough games for the effect of clawpomb, on higher TV games, to be important, even if it doesn't have as big an effect on the overall win ratio of various teams.
Further, considering we have major tournaments on this website, where regularly all the particpants will be 1500+ TV, then I think knowing how well teams do at a TV level that will enter a Major IS an important indicator of how well a team does, for THIS site. So we should probably have a third set of data for that TV range.
Please note: I purposefully tried to be vague about exact TV ranges, as I know that I am not expert enough to know exactly where the cut-offs should be. But the data is still there, and should still give us important information, which at the very least, will be very pertinent to THIS site, even if not to the world as a whole (although I think it will be important to the rest of the world, but I leave that as an argument for someone else.) |
|
|
Corvidius
Joined: Feb 15, 2011
|
| Posted:
Dec 23, 2011 - 11:26 |
|
Nelphine, i think you may have missed the point on this one. It's simply a list of tiers and the definition of why those tiers have been selected. ClawPOMB analysis is a different issue and i don't think we actually have a reliable source of data to analyse it as your tv ranges have to assume the presence of the combo or assume that it isn't present at lower levels, neither of these is a reliable assumption. |
|
|
koadah
Joined: Mar 30, 2005
|
| Posted:
Dec 23, 2011 - 11:49 |
|
|
Garion
Joined: Aug 19, 2009
|
| Posted:
Dec 23, 2011 - 12:06 |
|
I haven't read any of this thread, mainly because it was a two way conversation for the most part but now that has now calmed down (thank god).
So... is there anywhere I can see how specific races fair at different TVs?
To make a guess (just interested to see how close or far off i am) I would say the races would be ordered something like this at each level -
1000 - 1400
Woodelf
Amazon
Undead
Dwarf
Skaven
Pro Elves
Lizardmen
Orc
Chaos Dwarf
Human
Norse
Necromantic
Dark Elf
High Elf
Khemri
Chaos
Slann
Chaos Pact
Vampire
Nurgle
Goblin
Halfling
Underwold
Ogre
1401 - 1800
Skaven
Woodelf
Amazon
Dark Elf
Pro Elves
Necromantic
Lizardmen
Chaos Dwarf
Chaos
Dwarf
Slann
High Elf
Khemri
Nurgle
Chaos Pact
Norse
Undead
Human
Orc
Vampire
Underworld
Goblin
Halfling
Ogre
1801 +
Chaos Dwarf
Woodelf
Dark Elf
Chaos
Necromantic
Pro Elves
Skaven
Lizardmen
Amazon
High Elf
Nurgle
Chaos Pact
Khemri
Dwarf
Slann
Norse
Underworld
Undead
Human
Orc
Vampire
Goblin
Halfling
Ogre |
|
|
Sp00keh
Joined: Dec 06, 2011
|
| Posted:
Dec 23, 2011 - 12:47 |
|
here's my analysis
i took the raw data from hito,
i ignored who each game was played AGAINST
i retained the brackets but discarded results over 2210 for their lack of data quantity
i weighted each win percentage by the number of games played
so if lizards play 9 games against elves and win 8, and they lose 1 game against orcs, that means their win rate is 80%, rather than (88.8% + 0%) / 2 = 44%
my table is ranked by overall winningness.
the average win rate was 36% (draws counted the same as a loss)
pale green highlight on the main table is an above average result for that bracket
on the right there's 2 columns relating to trend.
i tracked the trend of change over time, so for example
lizards and undead get worse at higher ratings, (down arrow 'v' and red number)
chaos and the elf teams get stronger at higher ratings (up arrow '^' and green number)
this numerical trend change is the simple difference between the averages of the 960-1260-1460 brackets ('low') and the 1710-1960-2210 brackets ('high')
 http://i39.tinypic.com/vwsruu.png or http://img714.imageshack.us/img714/1231/newpicturee.png |
Last edited by Sp00keh on %b %23, %2011 - %13:%Dec; edited 1 time in total |
|
Sp00keh
Joined: Dec 06, 2011
|
| Posted:
Dec 23, 2011 - 12:51 |
|
to directly answer garion, my numbers show: (i overlapped the brackets when averaging for low/mid/high)
960-1460:
Wood Elf
Dark Elf
Amazon
Skaven
Elf
Chaos Dwarf
Lizardman
High Elf
Necromantic
Undead
Chaos
Human
Dwarf
Norse
Slann
Nurgle
Chaos Pact
Khemri
Orc
Vampire
Underworld
1210-1960
Wood Elf
Dark Elf
Skaven
Elf
Amazon
High Elf
Chaos Dwarf
Lizardman
Undead
Necromantic
Dwarf
Slann
Chaos
Chaos Pact
Norse
Human
Nurgle
Orc
Khemri
Vampire
Underworld
1710-2210
Dark Elf
Wood Elf
High Elf
Amazon
Elf
Skaven
Chaos
Necromantic
Chaos Dwarf
Nurgle
Lizardman
Vampire
Human
Slann
Undead
Underworld
Khemri
Dwarf
Norse
Chaos Pact
Orc |
|
|
Garion
Joined: Aug 19, 2009
|
| Posted:
Dec 23, 2011 - 12:56 |
|
cool thanks,  my predicitions were pretty accurate for the most part. Orcs are lower than i expected at a low Tv and chaos dwarves are lower than i expected everywhere which is a suprise. Apart from that the rest seem to be round about where i though
Its interesting how your stats put Wood elves at the top though and voodoo mikes make them bottom of T1 and top of tier 2. Maybe stats can be manipulated to suit your needs? |
_________________
 |
|
Vesikannu
Joined: Mar 06, 2011
|
| Posted:
Dec 23, 2011 - 12:59 |
|
| Garion wrote: | | Its interesting how your stats put Wood elves at the top though and voodoo mikes make them bottom of T1 and top of tier 2. Maybe stats can be manipulated to suit your needs? |
Well, Sp00keh counted draws as losses so that should have some effect on the results. |
|
|
JimmyFantastic
Joined: Feb 06, 2007
|
| Posted:
Dec 23, 2011 - 12:59 |
|
Pretty cool spookeh, maybe you could discount Ogres, Flings and Gobbos when working out the average win%?
They make nearly all races above average just cos they are so bad. |
_________________
Pull down the veil - actively bad for the hobby! |
|
Garion
Joined: Aug 19, 2009
|
| Posted:
Dec 23, 2011 - 13:02 |
|
I don't think Ogres, Flings and Gobbos were counted anywhere other than the first Tv bracket, they dont show on his reading, and in that TV bracket flings and goblins are okay really. Ogres are still pointless mind. |
|
|
|
|
| |
Learning BB in YouTu...
Search
