Wisden-CricViz Test MVPs of the Century

[Tibarn]

Wisden and Cricviz made a list of the top 30 Test cricketers of the current century, post 1st Jan 2000, in their magazine Wisden Cricket Monthly.

 

The rankings are based off of CricViz's model, here is an explanation of it 

Quote

WITH THE HELP OF CRICVIZ, THE WORLD’S LEADING CRICKET ANALYTICS COMPANY, WE REVEAL THE PLAYERS WHO’VE HAD THE MOST INFLUENCE ON TEST MATCHES SINCE THE TURN OF THE CENTURY

In simple terms, our match impact mode formulated by CricViz data scientist Sam Green, is a measure of how much better or worse a player has performed compared to the expected performance of an average player appearing in the same position in the same match. So, when we say Steve Smith has an average match impact of +43.6 that means he is worth +43.6 runs more than an average player playing in the same situation.

Cricket matches are won by scoring more runs than the opposition, so the model is built using runs as its measure. This is easy for batsmen, who operate in runs, but for bowlers it requires us to convert wickets into a runs value. This allows us to place batsmen and bowlers on the same scale. Top-order wickets are ‘worth’ more than lower-order wickets. relative value of performances in Test matches varies according to the nature of the match itself. For example, a century scored in a team innings of 600 is likely to be less valuable than a century scored in a team innings of 250.

The match impact model accounts for this by essentially working out how hard it is to bat and bowl in each innings of a match. This is based on the host country, venue and game data all being combined and weighted across the four innings of a match. This then produces an expected run environment for each innings of a match – the base level against which players are measured.

 

Quote

There are three important things to note about how the data is presented:
 

1) The overall findings of the model are presented in average match impact terms, rather than total match impact, because we are seeking to identify those players
who have had the most notable impact when they have  played, rather than those who have benefitted from playing a lot. We have used a minimum qualification of 30 Tests to remove players with a relatively insignificant body of work behind them.
 

2) The model tends to benefit those who are in the game more: primarily all-rounders but also bowlers more generally. This is simply because these players are given
more opportunity to impact a match.

3) It is important to stress that the findings of this model are not illustrative of who the ‘best’ player is; rather, they are a measure of which players have influenced the matches they have played in most significantly.
 

 

They ranked the top 30 players based off their model. I don't want to manually type all 30, so I will just post the top 5 and all the Indian players in the top 30. India has 5 players in the top 30, second to Australia which has 10. 

 

Top 5 Players in the Model

#1 MUTTIAH MURALITHARA   MVP RATING : 9 7. 5

#2 RAVI JADEJA                         MVP RATING : 9 7. 3

#3 STEVE SMITH                        MVP RATING : 9 1 .7

#4 GLENN MCGRATH                MVP RATING : 8 9 . 6

#5 SHAUN POLLOCK                 MVP RATING : 8 4 . 9
 

 

The 5 Indian Players in the Top 30

#2 RAVI JADEJA                              MVP RATING : 9 7. 3

#8 RAVICHANDRAN ASHWIN      MVP RATING : 8 3 . 9
#18 VIRAT KOHLI                           MVP RATING : 7 0 . 4
#25 VIRENDER SEHWAG               MVP RATING : 6 7. 8
#29 ANIL KUMBLE                         MVP RATING : 6 5 . 9
 

 

 

 

 

[Tibarn]

Here is the page for Ravi Jadeja 

 #2 Player of the century 

[Tibarn]

If one reads the blurb  about how CricViz's model works, one can see that they don't really factor in fielding into the equation. 

 

Muralidharan is #1 at a rating of 97.5

Jadeja is #2 at a rating of 97.3

 

If we factor in the fact that Jadeja is a next-level fielder, much better than Muralidharan, he would be the MVP of the century! 

Edited August 21, 2020 by Tibarn

[Gollum]

Does it take into account away performances? Cos I don't agree with the result. For me Steyn is the MVP of this century. 

[Tibarn]

9 minutes ago, Gollum said:

Does it take into account away performances? Cos I don't agree with the result. For me Steyn is the MVP of this century. 

Yes, they claim that their model accounts for home vs away. It also accounts for match situation, conditions, and quality of wickets(ie top order more valuable vs lower order). 

 

Steyn is actually ranked 27th in the top 30

They say that his average is deflated because of the bowling conditions in SA being favourable and that 32% of his wickets came from taking tail-end wickets. 

[Tibarn]

Cumulative MVPs of the decade. 

We have two of the top 10 here. Dravid at 8 and Sachin at 10.

[cricketfan28]

Sir!

[Ankit_sharma03]

In Indian condition in test format no doubt jaddu is the MVP....he ll be the 1st name in my XI in ATG XI india in Indian conditions

[zen]

A nice read but below is a bummer for me (based on what I have skimmed through):

 

Quote

In simple terms, our match impact mode formulated by CricViz data scientist Sam Green, is a measure of how much better or worse a player has performed compared to the expected performance of an average player appearing in the same position in the same match. 

 

If we look at Jadeja's regular batting positions and avg at home:

 

• #7 - 38 (11)
• #8 - 35 (11)

 

Expected performance and/or the players from the visiting team in the same match may not have such avg more often than not. Similarly for bowling where he does well in India. So Jadeja will get more "impact benefits"  

 

On the other hand, if you are a top order batsman, where the expected performance is relatively high, you would have to perform at a much higher level to probably match Jadeja's points

 

If you look at Jadeja's away record:

 

• Batting avg 30 -> he gets +19 for impact 
• Bowling avg 35 -> he gets +23 for impact 
• batting avg less bowling avg = -5

 

 

Quote

So, when we say Steve Smith has an average match impact of +43.6 that means he is worth +43.6 runs more than an average player playing in the same situation.

 

If S Smith bats at 7 or 8, where if the expected performance is calculated lower than 43.6, he would (at his current performance level) have a match impact of more than 43.6. It should be noted that Smith has achieved this impact level where expectations of performance are high  

 

 

In short, if expected performance at your batting/bowling position = minnow level for e.g. and you perform, you can get more "impact" points

Edited August 22, 2020 by zen