A most interesting piece of stat: Calling Varun

[Dhondy]

My God! If that table could speak, what stories we would hear!

[patriot]

My God! If that table could speak' date=' what stories we would hear![/quote'] Few things it hammers very strongly: 1) Gavaskar is MANY MANY notches above his peers. 2) Lara is a standout of his generation. 3)One must not doubt the veracity of this calculation seeing the likes of Younis Khan and Jayawardene featuring so highly. We know that they are poor performers away from subcontinent, therefore they can be safely overlooked.

[Guest BossBhai]

-- Removed on request of the user --

[zen]

Few things it hammers very strongly: 1) Gavaskar is MANY MANY notches above his peers. 2) Lara is a standout of his generation. 3)One must not doubt the veracity of this calculation seeing the likes of Younis Khan and Jayawardene featuring so highly. We know that they are poor performers away from subcontinent, therefore they can be safely overlooked.

What that tells me is that Gavaskar's team mates were not as good as say Richards or Chappells' team mates. Same goes for Lara vs Tendulkar/Ponting/Waugh/others To know who is the stand out player, you have to do a player vs player :winky:

[patriot]

So now....the top 10 from the period 1990 - Current ( excluding the bullies in their own backyard like..the Mahela's , Younis Khans, Moyos ..etc .) are: Lara: 9.2 Sangakarra: 8.03 Sehwag: 8.02 Anwar: 7.66 Kallis: 7.53 Ponting: 7.47 Pietersen: 7.44 Hayden: 7.35 Tendulkar: 7.29 Dravid: 7.20 Chanderpaul: 7.19 Above must be normalized with their respective sides bowling average during their career span.

[The Outsider]

And just for chuckles sake ... The % for Boss during the 90s was at 8.21 and peaked at 8.57 during 2002 ... after which Sehwag' date='RD and others started contributing heavily.[/quote'] Can you do a break up for Lara post/pre or with/without Walsh/Ambrose - that coupled with the above stat for Tendulkar should bear out the influence of a strong batting/bowling unit.

[Guest BossBhai]

-- Removed on request of the user --

[Dhondy]

Boss has unearthed golddust for Patriot. Look at the conclusions. 1. Era matters. Eleven of the top 13 are not from the modern era, and even Gavaskar has been long gone- nearly three decades. Why is it so? It's because of the phenomenon we have discussed thousands of times- the flattening out of pitches, making run scoring easier. But these are proportions of (both) team runs- that shouldn't vary even if run scoring is less onerous? It will. You see, because of the flat pitches, everybody will score runs, rather than the true greats. As a result, the proportion of team runs scored by the true greats will fall, although the difference will not be erased. In the pre-modern era, run scoring was difficult and the ones who shone emphasised the gulf between themselves and the also-batted. That's reflected in this table. 2. If you are from a weaker batting side, it obviously gives you an advantage. Why else would Gavaskar and Hanif appear ahead of Viv Richards? However, no way would that effect fully explain the gulf between Lara on one hand and Sachin and Dravid on the other. In fact, Ponting & Kallis appearing ahead of the Sachin-Dravid combo blows that theory out of the water. For the entire 90s, India struggled to find batsmen to complement Sachin, and it wasn't until 2001 that we had a world class batting line up. Wasn't it always said that Sachin carried the yoke of India's batting on his shoulders for the best part of a decade? Why is it not reflected in the table? I'll tell you why. Because Lara, besides his massive innings, also scored when others failed to score. Sachin's biggest knocks would not stand out in the same way. In any case, his percentage of 9.2, only behind Bradman, is simply unparalleled in the modern era, and cements his place as the greatest batsman this side of the 50s. 3.

[patriot]

Bowling averages for these batsmen's team during their playing career: Lara's WI- 33.23 Sangakarra's SL - 30. 04 Sehwag's India - 35.30 Tendulkar's India - 35.44 Ponting's Aus - 28.44 Now, how to normalize these with their % contribution in match ? Il come up with the remaining in a bit ...

[patriot]

Respective side's Bowling avg during playing career: Kallis' RSA - 29.97 Hayden's Aus- 28.14 Pietersen's Eng - 35.46 Saeed Anwar's Pak - 29.86 Dravid's India - 35.12

[Guest BossBhai]

-- Removed on request of the user --

[The Outsider]

A crude way would be to take some standard, like Tendulkar 35.44. So, on average India conceded 7 more runs per wicket than Australia during Tendulkar and Ponting's time. Ponting scored 11309 out of 151332 actual runs in 241 innings. Assuming the opposition batted same number of innings, if Ponting had Indian bowlers they would have conceded an extra 7 * 241 * 10 = 16870 runs. So he would have scored 11309 out of 168202(151332+16870) runs normalized to Tendulkar's bowling attack for a percentage of 6.7%

[patriot]

% contribution in match score ----------- Side's bowling average during career 1) Lara 9.2 ---- 33.23 2) Sehwag 8.02 ---- 35.30 3) Sangakarra 8.03 ---- 30.04 4) Anwar 7.66 ---- 29.86 5) Kallis 7.56 ---- 29.97 6)Ponting 7.47 ------ 28.44 7) Tendulkar 7.29 ------- 35.44 8) Hayden 7.35 ------ 28.14 9) Pietersen 7.44 -------- 35. 46 10) Dravid 7.20 -------- 35.12 How can the above be best normalized ? By simple multiplication of the 2 variables ???

[Lurker]

Btw, can you take that mysql-dump and import it into sql-server ?

Yeah right! From one dump to other dump :((:((

Lurker , those numbers in my previous post are correct ... unless you claim that SQL Server has bugs in addition

Man you should have put a large disclaimer suggesting this was all done on SQL server :giggle: Seriously though, just like I never seem to be in agreement with your views(aside from music) similarly I never have issues with your number crunching skills so no worries. I will put my response later, have to get my Month-end(and Quarter-end) reports in place for tomorrow. xxx

[patriot]

If we normalize by multiplying the player's respective sides bowling avg during their playing career to their contribution % of match total the rankings are as: Lara - 305.716 Sehwag - 283.106 Pietersen - 263.82 Tendulkar - 258.35 Dravid -252.86 Sangakarra - 241.22 Anwar - 228.72 Kallis- 226.57 Ponting- 212.44 Hayden - 206.82

[zen]

Boss has unearthed golddust for Patriot. Look at the conclusions. 2. If you are from a weaker batting side, it obviously gives you an advantage. Why else would Gavaskar and Hanif appear ahead of Viv Richards? However, no way would that effect fully explain the gulf between Lara on one hand and Sachin and Dravid on the other. In fact, Ponting & Kallis appearing ahead of the Sachin-Dravid combo blows that theory out of the water.

How much you score depends up on one or a combination of factors mentioned below: 1. Playing along with not that good team mates (comparitively) 2. Batting in the top order, which gives you a big chance to score big 100s when going is easy 3. Batting with an intent of winning a game or saving it 4. The difference b/w say 7.65 and 7.15 isn't huge. Anyone who says that someone with 7.65% of runs> someone with 7.15% runs by just looking at those stats (ignoring some of the other factors) and drawing a conclusion from that is probably smoking If those are the conclusions from that then this analysis was pretty much useless, imo

[Guest BossBhai]

-- Removed on request of the user --

[Dhondy]

But wouldn't an econometric regression be more effective here in accounting for far more deterministic/independent variables? If someone really had the time/effort to build one, it'd be possible to add in variables for just about every bowler (good or bad) and see the impact that the presence of said bowler had on the batsman's average (or whatever other y-variable you set).

We call it multivariate regression analysis, Thal. but it may not be necessary. Let's look at the variables that influence individual runs scored as a proportion of team totals, given that you are indeed a top-notch batsman. 1. Weakness of one's batting colleagues. The weaker they are, the higher will be the proportion of your runs. 2. The strength of the opposition's bowlers. The better they are, the less your mediocre colleagues will score in relation to you. OTOH, a weak opposition bowling would allow others to catch up with your scores. Everybody scored against Zim & BD, or against the India of yore. 3. Now, the corollary. Strength of your own bowling attack. The stronger they are, the less the opposition will score and therefore the more likely it is that your proportion of runs will be higher. 4. Similarly, great bats in the opposition will drag your proportion down. 5. Pitches. On flat pitches, everybody scores. Your proportion goes down. On tough pitches, you pi.ss on others. Thus, if a weak batting & bowling side is playing a strong batting & bowling side, say WI v Australia, you benefit from your weak batting colleagues and the opposition's strong bowling, but suffer because of your own weak bowling and the opposition's strong batting. The benefits are cancelled out by the disadvantages. That is why, Brian Lara would not benefit from WI playing Australia or SA. Next assume that a weak batting and bowling side is playing a side with strong batting but weak bowling, say WI v India.You (Lara) would again benefit from his own side's poor batting, but that single advantage is overwhelmed by your own weak bowling, the opposition's strong batting and the opposition's weak bowling, allowing your own mediocre batsmen to score freely. Next, let us take the example of a weak batting and bowling side, again the WI, playing a strong bowling, but a weak batting side, say Pakistan. In this instance, you benefit from your own batsmen's ineptitude (which is invariant), the opposition's bowling strength, and the opposition's weak batting. The disadvantage would be your own weak bowling, which lets the opposition score. Thus, in only one of the above three scenarios, does a batsman in a really weak side benefit. He loses out for one scenario and is neutral in the 3rd. You repeat the analysis with a side that has the polar opposite, say Australia, and you get exactly the same scenarios- one beneficial, one deleterious and one neutral. That's why I say, the table already factors in many of the influences that you think could make a difference. You'd be surprised how things even out, except of course the influence of flattening of pitches the world over in the modern era. You can't even that out, and it shows.

[The Outsider]

How can the above be best normalized ? By simple multiplication of the 2 variables ???

I've given one way to do it above. Simple multiplication would also be alright, but you will stand to lose the intuitive value of your result as it will no longer be in simple percentage terms.

[patriot]

If we further normalize the below stat - by using each batsmans team's batting avg during their playing career - it will perfectly sort out any of the flaws and give us a near perfect picture. Is that correct Dhondy/Shwetabh ?

If we normalize by multiplying the player's respective sides bowling avg during their playing career to their contribution % of match total the rankings are as: Lara - 305.716 Sehwag - 283.106 Pietersen - 263.82 Tendulkar - 258.35 Dravid -252.86 Sangakarra - 241.22 Anwar - 228.72 Kallis- 226.57 Ponting- 212.44 Hayden - 206.82