Statistics and Blood Bowl Leagues

[el sid]

I'm administrating the Turnhout Trasher League TTL website. This includes updating TD and CAS rates. It's a small league, but in a period of about 1 year we managed to play 94 games so far. A total of 30 teams have participated of which 21 remain active.

These numbers allow for some statistical analysis.

For a statistical analysis I isolated all teams which played at least 7 games. This results in 16 teams:

1 Skaven (15 games)
1 Lizardman (15 games)
1 Orc (14 games)
1 Chaos Nurgle (13 games)
1 Goblin (12 games)
1 Necromantic (11 games)
2 Dark Elf (both played 10 games each)
2 Dwarf (10 and 8 games)
1 Pro Elf (8 games)
1 Human (8 games)
1 Halfling (8 games)
1 Chaos Dwarf (7 games)
1 Ogre (7 games)
1 Amazon (7 games)

I believe this gives a nice, representative mix of official Blood Bowl teams. ST, AG and even the occasional fun team are mixed in the numbers.

The other 14 teams played 6 games or less. Arbitrarily I judged this too little to cancel out the effect of chance on the performance of the team.

Next I calculated normalized scores for each team:

- Ranking represents how well a team performs points-wise: How many games did each team win, draw and lose?
- Offence displays the offensive capacities of a team. This is determined by the amount of TD's scored by the team.
- Defence displays the defensive capacities of a team. This is determined by the amount of TD's a team allows to be scored by the opposition.
- Violence represents the CAS-causing capacity of each team. This is determined by the number of CAS's caused by the team.
- Robust shows how many CAS's the team received during the league.

Without going into detail, the scores show little surprises. Bash teams score good at Violence and Robust, AG teams excel in Offence.

The analysis is not yet finished, but I do came to following conlusions:

- Succes of a team is determined by good offence and defence. This can combine in different ways: Skaven teams tend to score a lot but also get a lot of TD's against. Dwarf teams tend to score less, but are very difficult to score against.
OK, this is no world-shocking news as the game is won by scoring TD's... But it is nice to see it confirmed in the numbers.

- Succes of a team is NOT determined by the amount of damage it causes or receives!
This I think is surprising. I would expect very violent teams to be succesful, while fragile teams should lose a lot. The numbers do not confirm this. Ranking scores, Violence scores and Robust scores tend to spread out very nicely, with no correlation visible.

This would mean succesful teams should focus on developing scoring and defensive players, instead of trying to win by destroying the opposition.

This is a very long post, I know. Thanks for staying with me this far. I was just wondering if anyone else tried something similar with his league data? Did you come to the same conclusions? Did you try something else?

I hope to extract more data later on.

el sid

[DoubleSkulls]

You may enjoy looking at this http://www.irwilliams.com/naf/

All the stats from every NAF game compiled by race.

[DamianTheLost]

Hmm haven't done any of the likes, but I like your enthusiasm. However, I find Blood bowl to be won on defence no matter the team in the long run. Scoring a lot eg. as WE won't win you games if you can't stop the opposition

[DoubleSkulls]

Generally the stats agree with you. The only race in the top 8 with a worse than average defence is skaven.

[stormmaster1]

TDs scored doesn't neccessarily reflect defence. I had a dwarf team only stop an opposition score on one kick off in 9 games. I still won 8 of them. Most bashy teams stall and score turn 8 so the opposition only recieves once. Most fast teams score every 2-3 turns so the opposition recieve many times per game. It would be useful to see whether more games are won from preventing the opposition scoring when they recieve, turning the ball over and scoring when they recieve or through good "clock management".