Elephant in the room

[akshayxyz]

Cricketers arranged by their innings per match. Notice Trescothik who has a higher innings /match count (1.88) has a runs/match of 77. However players like Hammond and Barrington who have superlative averages of 58 only have a runs/match of 83-85 because their innings/match is around 1.6. A difference of 15 points in average was reduced to 6 points after calculating the skewed measure. Its because players who get more chances to bat per match have higher values. This could be because their team is bats twice more often than other teams, or their batting positions are such that they are guaranteed to bat every time the team gets in. One more reason for deviations from the average is the ignoring of not outs.

Good work. Even after this, if some-certain-one tries to defend runs/test metric. Then I will have to do PhD in English to find appropriate words to express my feelings.

[zen]

Cricketers arranged by their innings per match. Notice Trescothik who has a higher innings /match count (1.88) has a runs/match of 77. However players like Hammond and Barrington who have superlative averages of 58 only have a runs/match of 83-85 because their innings/match is around 1.6. A difference of 15 points in average was reduced to 6 points after calculating the skewed measure. Its because players who get more chances to bat per match have higher values. This could be because their team is bats twice more often than other teams, or their batting positions are such that they are guaranteed to bat every time the team gets in. One more reason for deviations from the average is the ignoring of not outs.

Let's take Hayden, For the next 10 tests, on which would you bet your life's saving on from the two options: * Hayden scorring 1020 runs based on his avg of 51 or * Hayden scoring 840 runs based on his rpt of 84

[ganeshran]

Rett, you are making broad generalizations on an insufficiently sized sample. You are using two hypothetical scenarios to stress the validity of a metric that will be calculated from the performance in every match regardless of what the actual situation it is played in. In order to justify an illogical original premise, you are further making more illogical assertions.

[zen]

Rett, you are making broad generalizations on an insufficiently sized sample. You are using two hypothetical scenarios to stress the validity of a metric that will be calculated from the performance in every match regardless of what the actual situation it is played in. In order to justify an illogical original premise, you are further making more illogical assertions.

Would you please answer the question? 1020 or 840? Thanks

[ganeshran]

Would you please answer the question? 1020 or 840? Thanks

900 around. P.S: Your justification of runs per test is still illogical.

[zen]

900 around. P.S: Your justification of runs per test is still illogical.

Thanks ... that's closer to the rpt case closed

[The Outsider]

Why not averages? - Usually inflated due to not outs - does not depict the conditions encountered, the team strength properly

How are they inflated by not outs in the case when the batsman has already played a substantial knock without being dismissed, something like 40-50+ chasing a 200+ score etc.? Lost at the second point.

For example, let's take two venues: durban and ahmedabad and assume that player A plays most of his games at Durban, player B plays most of his game at A'bad. Assuming both the players have the same skills, who do you think is likely to avg more?

The one at Durban. However, there are no real examples of this. Modern day players play their cricket on a variety of pitches, most batting friendly some not. What about getting familiar with conditions? It's not that South Africa hosts exclusively bowling friendly tracks - what if the weight of 2 bowling friendly tracks like Perth, Brisbane, Barbados, Sabina Park is outweighed or equaled by a doubly batting friendly track like Adelaide and Antigua?

At Durban: bowlers would be on top. What does that mean? The chances of a team getting bowled out are more. Which means that the player will get to play 2 innings. He has to be prepared for that. At A'bad: batsman would be on the top. The chances of getting a huge score are high with a declaration on cards. How does a player go in to this game? I guess, prepared to bat once, get a big score. Try to put the opposition under pressue. And ofc, if you are batting 2nd, try to bat as much as you can (assuming the other team has a good first inning score) Team strength: Let's assume player A's team has a weak batting line up, Player B's team has a strong batting line up How does player A condition himself? He is prepared to do more hard work. Usually comes in earlier than he should and thus more chances of facing fresh bowlers. Because his team is weak, more often he will have to be streched out. How does playeB condition himself? He knows he has to score big in the opportunity he gets. At times, he is walking in after a good start and milking the bowling. Sometimes, he comes in at 200/2 and smashes the tired bowlers to get an unbeated 100 as well A few more examples: Player A scores 55 and 35 at Durban, avg 45 Player B scores 70 and 10* at A'bad, avg 80 Both have scored 80 runs in the game in the enviornment that they are presented with. Player A scored 80 runs in tougher condtions. Player B scored a 70 in a score of 450/7 dec and probably batted a little in a draw to end up with a 10 not out

But the fact is that Tendulkar and Dravid average more(or roughly so) away as at home - clearly negating the above theory of taking advantage of "dead home tracks" to boost their averages. Whereas the likes of Ponting and Lara average considerably less abroad than at hom. Comments would be welcome at this point?

Let's say a guy avg 50 and has rpt of 80. In the next 10 tests, would you bet (if you had to on the option given to you) on him scoring 1000 runs (based on his avg) in the next 10 tests based on his avg or 800 runs based on rpt. To me him scoring at 80 rpt looks more likely .... These are just a few examples.

Of course, you wouldn't bet on the actual realization of the mean above the SD!

By taking runs per tests, lets call it rpt, I am attempting to capture "the enviornment". The more tests, you play, rpt captures even more diversity.

You're contradicting yourself - so, Tendulkar has played more tests than Lara and has encountered more "diversity"? Is that what you are saying?

Is rpt 'fool-proof'? well, nothing is. So I am not claiming it is a perfect system. Only that it captures better the essense of what I am trying do, i.e. mainly forecast I am only explaining rpt, not opening it up for discussion or someone telling me whether I should use it or not.

No, it doesn't capture any essence as outlined above. If you still want to use it, good for you.

[zep1706]

I don't understand why on earth the great man still gets flustered nearing major milestones. After scoring so many international centuries he still gets terribly nervous in the 90s. I wish he gets it before the Aus tour. We can do without the drama in Australia.

[ganeshran]

Thanks ... that's closer to the rpt case closed

:facepalm: You do know how averages work, dont you? It gives preference to not outs. So if Matty hayden (or any other player) stays not out even 2 times in 10 matches, his average will be higher than the indicative runs scored in those test matches. Runs per test treats out and not out exactly the same way. Like I said before, you are making further illogical assertions to justify an original premise which in itself makes no sense.

[Tendya]

@Rett : RPT makes no sense at all, let me put forward 3 points 1. I will pick 2 names with the same rpt , say Laxman and Atherton who both have rpt of 67, ( I am not bringing subjective factors like how Laxman was good in tough conditions or how Atherton sucked against genuine good bowling) Laxman averages above 47 while Atherton averages 37, i.e 30% more than Atherton. So as per rpt , they are equal. :hysterical: 2. Also capturing essence is all bull ****, runs per matches is directly dependent on the number of innings one gets to play, if there is a mass exodus of good players(like Aus in mid 80s, Aus now, or WI in late 90s) one gets to play more and his rpt would increase. Lara was in a comparitively strong team till the 2000 (Ambrose, Hooper etc) , he had a rpt of 82, in the next due to the more innings he played as he was in a weak team (and also due to the strong form i should acknowledge), his rpt shot upto 91 3. Also your argument that it reflects the number of runs they score int the next few matches is totally wrong. Lets take Richards now he used to have a rpt of 75, but his rpt dropped to 55 during his last phase, if he went on 2 play 22 years like Sachin, his average would have fallen to the low 40s And No Sir, you are not trying to bring sanity by educating Sachin fanatics, you are just making fool of yourselves

[zen]

:facepalm: You do know how averages work, dont you? It gives preference to not outs. So if Matty hayden (or any other player) stays not out even 2 times in 10 matches, his average will be higher than the indicative runs scored in those test matches. Runs per test treats out and not out exactly the same way. Like I said before, you are making further illogical assertions to justify an original premise which in itself makes no sense.

When I am trying to "forecast", averages don't help me. So I have to rely on something like rpt, unless you have something better to suggest So if I want to now how much a player can score (hypothetically), if he had played 180 tests. I would be better served by using rpt than avgs

[zen]

@Rett : RPT makes no sense at all, let me put forward 3 points 1. I will pick 2 names with the same rpt , say Laxman and Atherton who both have rpt of 67, ( I am not bringing subjective factors like how Laxman was good in tough conditions or how Atherton sucked against genuine good bowling) Laxman averages above 47 while Atherton averages 37, i.e 30% more than Atherton. So as per rpt , they are equal. :hysterical: 2. Also capturing essence is all bull ****, runs per matches is directly dependent on the number of innings one gets to play, if there is a mass exodus of good players(like Aus in mid 80s, Aus now, or WI in late 90s) one gets to play more and his rpt would increase. Lara was in a comparitively strong team till the 2000 (Ambrose, Hooper etc) , he had a rpt of 82, in the next due to the more innings he played as he was in a weak team (and also due to the strong form i should acknowledge), his rpt shot upto 91 3. Also your argument that it reflects the number of runs they score int the next few matches is totally wrong. Lets take Richards now he used to have a rpt of 75, but his rpt dropped to 55 during his last phase, if he went on 2 play 22 years like Sachin, his average would have fallen to the low 40s And No Sir, you are not trying to bring sanity by educating Sachin fanatics, you are just making fool of yourselves

rpt is useful as a forecasting tool duh .... so even though Lax is better than arthoton, rpt will tell you that he would score as many as Laxman if they were to play x amt of tests. already explained rpt so wouldn't go into that again

[Tendya]

rpt is useful as a forecasting tool duh .... so even though Lax is better than arthoton, rpt will tell you that he would score as many as Laxman if they were to play x amt of tests. already explained rpt so wouldn't go into that again

You have used RPT to compare Lara and Sachin, i will like to use it to compare Atherton and Laxman and say both are equally good, fair enough right :two_thumbs_up:

[Yoda-esque]

You have used RPT to compare Lara and Sachin' date=' i will like to use it to compare Atherton and Laxman and say both are equally good, fair enough right :two_thumbs_up:[/quote'] Stop wasting your time!!!

[zen]

You have used RPT to compare Lara and Sachin' date=' i will like to use it to compare Atherton and Laxman and say both are equally good, fair enough right :two_thumbs_up:[/quote'] I said that Lara would have scored more if he had played 180 tests. And yes, I based my forecast on rpt Why Lara is a better test player than Ten, I gave my reasons in one of the posts.

[zen]

How are they inflated by not outs in the case when the batsman has already played a substantial knock without being dismissed, something like 40-50+ chasing a 200+ score etc.? In either case, you would be better of forecasting based on rpt than avg. Say he scored a 50 too in the first inning, you are still better of forecasting runs based on 100 runs per tests, than 100 runs per inning (200 runs per test) Lost at the second point. The one at Durban. However, there are no real examples of this. Modern day players play their cricket on a variety of pitches, most batting friendly some not. What about getting familiar with conditions? It's not that South Africa hosts exclusively bowling friendly tracks - what if the weight of 2 bowling friendly tracks like Perth, Brisbane, Barbados, Sabina Park is outweighed or equaled by a doubly batting friendly track like Adelaide and Antigua? But the fact is that Tendulkar and Dravid average more(or roughly so) away as at home - clearly negating the above theory of taking advantage of "dead home tracks" to boost their averages. Whereas the likes of Ponting and Lara average considerably less abroad than at hom. Comments would be welcome at this point? I said in my post that these are only a few examples, there are tons of scenatios but I need a simple system which can help me to forecast runs, which is where rpt comes in Of course, you wouldn't bet on the actual realization of the mean above the SD! You're contradicting yourself - so, Tendulkar has played more tests than Lara and has encountered more "diversity"? Is that what you are saying? What I am trying to say is that if someone has played a considerable amount of tests, rpt tends to capture the enviornment he usually encounters better (not with laser precision though) No, it doesn't capture any essence as outlined above. If you still want to use it, good for you.

responses in "blue"

[Magneto]

Seems like those statistics classes that I bunked ages ago - probability, permutation, combination, rpt, averages, median what not?? Ufff!!!

[ganeshran]

If Andy Ganteaume had played 180 tests he would have scored 20160 runs. Legend Alas, if only data could be extrapolated with such ease.

[zen]

If Andy Ganteaume had played 180 tests he would have scored 20160 runs. Legend Alas, if only data could be extrapolated with such ease.

If you need a complicated system to do that be my guest :winky: Obvioulsy, there are various factors but when we talk about proven performers, it's not difficult to say, for example, if Gavaskar had played 150 tests, he would have got 12k runs Those who don't like Gavaskar would probably say what if he failed to go past 50 runs in any test after is 125th test. That's a probablilty too .... But we have our senses to know which is more probable, i.e. Gavaskar reaching 12k runs

[akshayxyz]

Seems like those statistics classes that I bunked ages ago - probability' date=' permutation, combination, [b']rpt, averages, median what not?? Ufff!!!

no, not 'rpt'.