Is he 'Test Class'? Mediocre? Fast Bowling Edition

[silentstriker]

First, we can say with relative certainty that Mohammad Sami and Ajit Agarkar are quite possibly even bigger statistical anomalies than even Bradman.

The question is often asked if someone is 'mediocre', or if they are even 'Test class'. I decided to look at all the fast bowlers who have played since the turn of the century, and try to figure that out.

I took all fast bowlers who have taken thirty wickets in their careers, and plotted them on a normal distribution:

The black dots are the actual figures of bowlers (again, 30+ wickets, Bangladesh/Zim removed).

The average is approximately 31, and the standard deviation is 5. Basically, for you to be an 'average' bowler in this decade, you have to average around the 31 mark total.

2SD+: Exceptional Performance
1SD+: Great Performance
0.5SD+: Above Average Performance
-0.5-0.5SD: Average Performance
-0.5- SD: Below Average Performance
-1SD: Bad Performance
-2SD: Terrible Performance
-3SD: You can't win a Test
-4SD: WTF?
-5SD: Ajit Agarkar Standard

For reference, by the time you get to our 'WTF?' standard, you are in the 99.997th percentile of averages. Meaning, on average, if three thousand bowlers play Test cricket, you're #3000 in averages. I can't bring myself to categorize the great Ajit Agarkar Standard.

Note: If minimum of wickets were increased to 50, Ambrose, Cork, and a few others would not be on this list.

If we go by the 30+ criteria:

Exceptional Performers

1. Ambrose
2. Walsh
3. McGrath

Great Performers

1. Bond
2. Akhtar
3. Ahmed
4. Steyn
5. Asif
6. Clark
7. Cork
8. Pollock
9. Donald

Above Average Performers

1. Gillespie
2. Gough
3. Younis
4. Bichel
5. Jones
6. Sidebottom
7. Johnson
8. Ntini

Average Performers

1. Cairns
2. Lawson
3. Vaas
4. Hayward
5. Akram
6. Caddick
7. Hoggard
8. Hilfenhaus
9. Tuffey
10. Srinath
11. Lee
12. Colleymore
13. Siddle
14. King
15. Kaprowicz
16. Harmison
17. Nel
18. Cuffy
19. Sreesanth
20. Kallis
21. Morkel
22. Pathan
23. Flintoff
24. Franklin
25. Khan
26. Mills
27. Oram
28. Dillon
29. O'Brien
30. Martin
31. Collins
32. Malinga

Below Average Performers

1. Sharma
2. Gul
3. Anderson
4. Zoysa
5. Fernando
6. Broad
7. Taylor
8. Hall
9. Patel

Bad Performers

1. Razzaq
2. Bravo
3. Singh
4. Edwards
5. Nehra
6. Drakes

Terrible Performers

1. Sanford

How do you win a Test?

1. Powell

WTF?

1. Sami

Ajit Agarkar

1. Agarkar

Now, what about for a specific country? You may be wondering what the 'average' performance would be for a pace bowler playing in India? Using the methodology that you must have taken at least eight wickets, we can come up with an idea. Now, the standard deviation here is much higher (but the sample is much smaller), so I'm still going to use the 5 point marks to differentiate the bowlers. An average fast bowler should average around 36 in India (so the "average" fast bowler averages 31 overall and 36 in India).

But for India, we have bowlers who average 15points lower than the average pace bowler in India (e.g, average less than 21!) For that, we have to introduce a new 'Can I borrow your genes?' category.

Can I borrow your genes:

1. Steyn
2. McGrath

Exceptional

1. Gillespie
2. Hoggard
3. Dillon

Great

1. Pollock
2. Ntini
3. Anderson
4. Sharma
5. Flintoff
6. Srinath

Above Average

1. Watson
2. Akhtar
3. Patel

Average

1. Sreesanth
2. Khan
3. Morkel

Below Average

1. Kasprowicz
2. Johnson

Bad

1. Balaji

Terrible

1. Streak

WTF?

1. Pathan

Ajit Agarkar
None

Unfortunately, we're not done yet. We have to introduce categories even below Ajit Agarkar, and failure this bad has no word for it in the English language.

........

1. Lee
2. Agarkar
3. Sami (14 wickets with almost 1000 runs conceded...average of close to 70!)

 

[marc71178]

Hmm, any rating that has Cork as great is wrong.

 

[silentstriker]

36 wickets, he would not be here if the wicket criteria was 50. He had one great series where he took 20 wickets @ 12.25, which pushes him high. The others were terrible, and he was dropped four series later.

 

[Top_Cat]

Questionable assumption of normality.

 

[L Trumper]

Good analysis.
Try the same for strikerates also. Clubbing them together might give better idea about bowlers (Statstically).

 

[Craig]

 

[silentstriker]

Questionable assumption of normality.

Shh... Everything is normal. It must be.

It's like in Physical Chem, everything must have y=mx+b applied to it. No matter what.

 

[silentstriker]

Good analysis.
Try the same for strikerates also. Clubbing them together might give better idea about bowlers (Statstically).

I don't necessarily think it would really change anything here.

 

[AaronK]

How did wasim akram end up as an average performer?

The greatest bowler i have seen..as good as McGrath, Ambrose, Walsh, and even Marshal..

 

[silentstriker]

How did wasim akram end up as an average performer?

The greatest bowler i have seen..as good as McGrath, Ambrose, Walsh, and even Marshal..

It's from 2000 onwards. He only played 13 Tests after 2000, and took only 31 wickets at an average of 30 in his last two years.

 

[bryce]

Exceptional

1. Dillon

Great

1. Pollock
2. Ntini
3. Anderson
4. Sharma
5. Flintoff
6. Srinath

Enjoyed reading that part tbh. Yeah i guess the wicket criteria should have been 50

 

[Dano.85]

Well done that must have taken ages to do. But i think i will stick to my eyes and my opinion who i think is good or not, while i honestly think what you have done is good great bowlers have so much more then good stats, they are larger then life, they are leaders, they dont follow a plan they make it, it might not all ways go right thus a few bad games/series stats wise. But no maths, graph or statistical analysis take that into account.

 

[shortofalength]

How can Mervyn Dillion make a list of bowlers as exceptional and Jimmy Anderson great, just shows how limited a stastically analysis is?

 

[morgieb]

How can Mervyn Dillion make a list of bowlers as exceptional and Jimmy Anderson great, just shows how limited a stastically analysis is?

Dillon & Anderson are 'great' and 'exceptional' because they got good records in India, ITBT. Aren't that good elsewhere.

 

[Richard]

This'd be so much more interesting if you started in September 2001 rather than January 2000.

 

[G.I.Joe]

Al Qaeda has nothing to do with this.

 

[Richard]

I know. Sadly there's so many more significant dates concerning Al Qaeda than just 11th September 2001, they just tended to get lesser coverage.

 

[The_roc]

The lower you pitch the qualifying standard the more likely you are to get rogue results.

Wonder what the stats would look like if you used 100 wickets as the cut off.

 

[irfan]

Interesting analysis SS. Reckon 75 wickets would be a good cut-off. Reckon you should multiply the bowling S/R with their average and normalise that - because I reckon quick strikers like Waqar are hard done by if you only consider their average