Official Lance Armstrong Thread: Part 3 (Post-Confession)

[thehog]

I thought insurance companies were supposed to be experts at statistics. I don’t understand their reasoning here. I assume the bonus reflects their estimate of the probability or difficulty of achieving various levels. They award $1.5 million for winning in 2001 and 2002, or $750,000 per win. But if he wins again in 2003, he gets $3 million, which means $1.5 million for winning a third time. Then $5 million for winning four years in a row, which means $2 million for winning the fourth time.

IOW, they are betting that after 2002, each TDF victory is less likely than the year before. But that doesn’t make sense. A prior win should make a succeeding win more likely, because it establishes that you have what it takes to win. You’re betting on the basis of the information you have, and if the rider has managed to win once or twice in a row, that information says a third win and a fourth win are more likely, not less likely. It’s like flipping a coin, except that every time the coin comes up heads, it’s somewhat more likely to come up heads the next time.

Maybe they thought because LA was 27 when he won his first, that his form would be in decline by 2003, when he was 31? I guess I could see that, most multiple GT winners are finished in their early 30s. But as was discussed at the time, having been out of racing for a couple of years because of the cancer, and on a relatively light schedule that was focussed on peaking at the TDF, LA was poised to go on somewhat longer than his predecessors.

I won’t even comment on all the typos in the email, and don’t say “it’s only an email”. I would be ashamed to make that many mistakes on a post in an internet forum, let alone a business communication.

Gamblers fallacy?

It does make some sense from the perspective if their chance algo in excel was based on 5 wins and breaking that barrier. The likelihood based on historical data was unlikely. Indurian topped out at 5 etc. So once Armstrong hit it in 2003 the likelyhood of winning in 2004 and 05 became less (based on the history of the Tour).

Bobs son was the guy who put together the math. Poorly done and stupid to even accept Armstrong's money in the first place. Drug use clearly not accounted for!

Gamblers remorse!

 

[KingsMountain]

I thought insurance companies were supposed to be experts at statistics. I don’t understand their reasoning here. I assume the bonus reflects their estimate of the probability or difficulty of achieving various levels. They award $1.5 million for winning in 2001 and 2002, or $750,000 per win. But if he wins again in 2003, he gets $3 million, which means $1.5 million for winning a third time. Then $5 million for winning four years in a row, which means $2 million for winning the fourth time.

IOW, they are betting that after 2002, each TDF victory is less likely than the year before. But that doesn’t make sense. A prior win should make a succeeding win more likely, because it establishes that you have what it takes to win. You’re betting on the basis of the information you have, and if the rider has managed to win once or twice in a row, that information says a third win and a fourth win are more likely, not less likely. It’s like flipping a coin, except that every time the coin comes up heads, it’s somewhat more likely to come up heads the next time.
<snip>

But this isn't like "flipping a coin", because the odds of a positive outcome for the rider are much less than 50%. Let's assume that the odds of Armstrong winning a single TdF is 1/3. Actually if the insurers thought the odds were that high, they would have asked for a bigger premium. But anyway working with 1/3 and naively assuming for the moment that the odds of winning don't change with each win, the chances of the first 2 wins would be 1/9, the 3rd would be 1/27 and the 4th would be worth 1/81. Under this assumption, the payout for the 4th win should be 18 times the average payout for the first 2 wins.

Now if we believe that the odds of winning the 4th time improve to, say, 1/54, it is still worth 12 times the average of the payout for the first 2 wins. There is no reason at all to value the 4th win equally with the first win.

 

[ChewbaccaD]

I think we can all agree that SCA deserves their money back. They insured for Tour wins, and Lance never won the Tour, so I think any reasonable person would see that there has been an ill-gotten gain that needs to be returned. If only Lance had the honesty and moral fortitude to do what most reasonable people would do.

 

[DirtyWorks]

LA’s best hope here may not be the legal arguments for and against, but the sheer incompetence government lawyers sometimes display.

In the U.S., it's not incompetence it's some combination of
-not illegal
-not enforceable
-weak/non-existent penalty
-politically unpopular until it gets absurd.

The interested parties in keeping anti-doping out of the judicial system is a who's-who of sports:
NCAA (aka BCS)
NBA
MLB
USOC and IOC
PGA (the U.S. Olympic federation)
... and more

On top of that, we know most fans are okay with doping. As long as it's not their kid, it seems like it's fair game.

 

[D-Queued]

I thought insurance companies were supposed to be experts at statistics. ...

A prior win should make a succeeding win more likely, because it establishes that you have what it takes to win. You’re betting on the basis of the information you have, and if the rider has managed to win once or twice in a row, that information says a third win and a fourth win are more likely, not less likely. It’s like flipping a coin, except that every time the coin comes up heads, it’s somewhat more likely to come up heads the next time.

...

I won’t even comment on all the typos in the email, and don’t say “it’s only an email”. I would be ashamed to make that many mistakes on a post in an internet forum, let alone a business communication.

Agreed on the mistakes. Notably, the email to the UCI was from Lance's representatives.

WRT likelihood of winning, agree completely. You can also add the 'Patron' phenomenon, whereby successive wins come easier because the Peloton expects and adapts for it.

We did discuss this extensively some years back (and I know you were part of that dialog).

In the end, Chewie is right. Lance didn't win anything.

I think we can all agree that SCA deserves their money back. They insured for Tour wins, and Lance never won the Tour, so I think any reasonable person would see that there has been an ill-gotten gain that needs to be returned. If only Lance had the honesty and moral fortitude to do what most reasonable people would do.

...
Good post Race

Good catch Dave.

Dave.

 

[thehog]

In the end, Chewie is right. Lance didn't win anything.

I wonder what the payout was on no winner?

 

[Race Radio]

I think we can all agree that SCA deserves their money back. They insured for Tour wins, and Lance never won the Tour, so I think any reasonable person would see that there has been an ill-gotten gain that needs to be returned. If only Lance had the honesty and moral fortitude to do what most reasonable people would do.

But, but, lance is sorry. Really sorry. I know he is because he said it in all the interviews he has been doing lately.

 

[thehog]

I think we can all agree that SCA deserves their money back. They insured for Tour wins, and Lance never won the Tour, so I think any reasonable person would see that there has been an ill-gotten gain that needs to be returned. If only Lance had the honesty and moral fortitude to do what most reasonable people would do.

I just hope AIG get their money back. They need it. Perhaps a fund set up for tax payers to contribute?

Although I'm not sure on the ill gotten part. It's been well proven here that SCA did a lot of the "ill gotten" themselves. Sounds a lot like buyers remorse to me.

We'll just have to wait and see what the full story is. We tend to be finding out new bits of information everyday.

 

[D-Queued]

But, but, lance is sorry. Really sorry. I know he is because he said it in all the interviews he has been doing lately.

He has been trying really, really hard. But, has he been able to make himself cry yet?

Maybe he should put some onions in a zip lock, or something.

Dave.

 

[ChewbaccaD]

I just hope AIG get their money back. They need it. Perhaps a fund set up for tax payers to contribute?

Although I'm not sure on the ill gotten part. It's been well proven here that SCA did a lot of the "ill gotten" themselves. Sounds a lot like buyers remorse to me.

We'll just have to wait and see what the full story is. We tend to be finding out new bits of information everyday.

Good to see that wins procured from doping are not a problem for you. I'll send you your Chris Froome t-shirt tomorrow! I mean, anyone who sees things from your perspective can surely appreciate Froome's style.

 

[thehog]

It's a misstep, but not terribly uncommon in a case with this much information being exchanged. Lance's attorneys are not supposed to use it once informed, but...

Giving up witness interviews?

It’s unclear exactly what the documents contain — much of the litigation is sealed — but court documents suggest they include memoranda summarizing government witness interviews

That doesn't appear too common.

I guess it's a strategic misstep more than anything.

 

[Race Radio]

Giving up witness interviews?

That doesn't appear too common.

I guess it's a strategic misstep more than anything.

Yes, it is. They would have to share them eventually.

 

[mewmewmew13]

Maybe it was an 'intentional leak' to help speed **** up.
Ghod knows that anything to move it along would be positive...or maybe Lance's team will inadvertently let something slip w/r to their planned strategy.

 

[Race Radio]

Good to see that wins procured from doping are not a problem for you. I'll send you your Chris Froome t-shirt tomorrow! I mean, anyone who sees things from your perspective can surely appreciate Froome's style.

Lance thinks wins procured from doping are wrong too. That is why he settled with Acceptance Insurance, the folks who paid on his first Tour "Victories". They got a big check.

With SCA it is personal......but he is really, really, sorry.

 

[ChewbaccaD]

Giving up witness interviews?



That doesn't appear too common.

I guess it's a strategic misstep more than anything.

Giving up any work product is bad...but you don't understand that I guess?

Anyway, with huge batches of documents to be assessed, sometimes things like that happen. Not completely uncommon, but again, I can see how someone like you wouldn't know that. No biggie.

 

[ChewbaccaD]

Lance thinks wins procured from doping are wrong too. That is why he settled with Acceptance Insurance, the folks who paid on his first Tour "Victories". They got a big check.

With SCA it is personal......but he is really, really, sorry.

Yea, and some people are willing to let petty insecurities and a need to find like minded people to determine their allegiances. Pretty common among people of weak moral character. It isn't a surprise at all when they gravitate toward someone like Armstrong. He is the embodiment of their Gestalt, so it's like flies to sh...well, I'm really trying to be less colorful here so I'll just leave it there.

Anyway, Wonderboy has a lot of trouble on his hands...he's getting slaughtered financially, but I'm sure his attorneys don't blink a second when he tells them to keep fighting. He is the definition of job security for an attorney who is on retainer for him. Dang, lucky guys.

 

[Berzin]

Funny thing is these allegedly great legal minds (the best money can rent?) aren't coming up with any brilliant legal strategies commiserate with their billable hours.

Judging from the link Race Radio posted, they are using the type of low-end stalling tactics one would expect from a lawyer like Saul Goodman.

 

[Merckx index]

But this isn't like "flipping a coin", because the odds of a positive outcome for the rider are much less than 50%. Let's assume that the odds of Armstrong winning a single TdF is 1/3. Actually if the insurers thought the odds were that high, they would have asked for a bigger premium. But anyway working with 1/3 and naively assuming for the moment that the odds of winning don't change with each win, the chances of the first 2 wins would be 1/9, the 3rd would be 1/27 and the 4th would be worth 1/81. Under this assumption, the payout for the 4th win should be 18 times the average payout for the first 2 wins.

Now if we believe that the odds of winning the 4th time improve to, say, 1/54, it is still worth 12 times the average of the payout for the first 2 wins. There is no reason at all to value the 4th win equally with the first win.

Thanks for this post. It got me thinking, and I realize my criticism of SCA was off, though I believe your argument is, too. Here’s my reasoning.

First, the fact that the odds of winning are less than 50% (if they indeed were) is irrelevant to the argument. The argument you’re making would be the same for a 1 in 2 chance of winning once.

Second, the correctness of your argument depends on exactly how the deal or “bet” was structured. You’re assuming that it’s all or none. If at any time during the four year period LA fails to win a TDF, then he loses. If that is the case, then, yes, the payout for four in a row should be a great deal larger.

E.g., if the odds of winning two in a row are 1 in 9, and the payout is $1.5 million, then the premium should be $187,500. This is because the odds of losing are 8 in 9, and $187,500 x 8 = $1.5 million (neglecting profit margin, etc). Continuing with your example, if the odds of winning two in a row are 1 in 9, the odds of winning four in a row are 1 in 81. The odds of losing are 80 in 81, so if the same premium of $187,500 is paid, the payout should be $187,500 x 80 = $15 million.

This is correct if LA was given a choice, go for two in a row, three in a row, or four in a row. But my understanding—and someone correct me if I’m wrong—is that it wasn’t all or none. He was in the game for all three possibilities or levels. If he won two in a row, then he collected $1.5 million, even if he didn’t win three in row. If he did win three in a row, then he collected $3 million, even if he didn’t win the fourth. IOW, all three levels listed in that post were operative. Was that not the case?

Assuming this was the case, it changes the situation considerably. Let’s say he wins the first two. He or someone on his behalf paid the $187,500 premium, and won on the 1 in 9 odds, so the $1.5 million payout is just as it should be. At this point, the odds of winning the next one, making it three in a row, are 1 in 3. But since he already has paid the premium, and won that level 1 bet, anything more he makes for winning three in a row would be a no-lose proposition.

So the initial premium has to be revised to take this into account. It has to be more than $187,500, because if the 1 in 9 possibility does occur, then there is a 1 in 3 possibility of winning more money. Again, the odds are 8 in 9 he won’t win two in a row, but there is a 1 in 9 chance he will, and a 1 in 27 chance he will win three. If he wins $3 million for three in a row, then there are 24 chances in 27 he won’t win two in row and $1.5 million, 2 chances in 27 he will win only $1.5 million, and 1 chance in 27 he will win $3 million. So the premium should be [($3 million x 1) + ($1.5 million x 2)]/24 = $250,000.

We can do the same calculation for four in a row, with a payout of $5 million. If I have done this right, it comes out to $277,778.

So to summarize, even if we assume the same odds for winning each TDF, the payouts for winning the third or the fourth can be larger than those for winning the first and second. The premium can be adjusted to take this into account. And the payouts for these later wins do not reflect the very high odds against getting this far, because at earlier points, LA could collect for fewer wins.

Edit: The premium for winning two in a row, as initially calculated, should be $1.5 million/9, not 8, or $167,000. For three in a row, revised to $6 million/27, or $211,000, with four in a row possible, $247,000. My calculations earlier were assuming the payoff was a net above the premium, unlikely to be the case.

 

[Granville57]

How can one derail a thread about Armstrong, talking about Armstrong?

True. It's never been done before. And for the life of me, I can't imagine that it's even possible.

Hey, what happened to Hoggie?

 

[KingsMountain]

MI,
The calculation I provided does not depend on an "all or nothing" concept. Assuming invariant 1/3 odds of winning 1 TDF, the odds of winning the first two are 1/9, the odds of winning the first 3 are 1/27 and the odds of winning the first 4 are 1/81.

There are three separate bets here, and provided the bets are made before the first race, the a priori probablities are unchanged by what actually may occur in the races. The relative sizes of the bonuses don't need to be related; all that matters to the insurers is that they price each of the three bets according to the proposed bonus scheme. The deal could have been structured to pay 900K for the first 2, $2.7M for the first 3 wins, and 8.1M for the 4th win. In that case, each bet would need a premium of $100K.

Or they could have structured the deal to give a payouts of $900K for each of the three bets. In that case the first bet still needs a $100K premium, but the second bet needs $33K, and the third bet needs $11K, for a total of $144.5K. The actual deal was structured intermediate to the two above examples.

(Again, in real life, in this kind of situation the insurers only have a rough idea of the true odds, and they will price the premiums corresponding to odds which are perhaps 2 or 3 times higher than their estimate. SCA did not believe that the odds of Armstrong winning a single TdF were 1/3.)

 

[Merckx index]

I don't disagree with this. As I implied in my previous post, the payout can be anything, if the premium is adjusted accordingly. And you have done this in this example of yours. If one takes on all three bets, then the premium is tripled over what it would be if one only took the first bet.

I was responding to your original post, when you said the payout for four wins should be eighteen times that for one win, without specifying that the premium would have to be much higher than it would be on just winning two in a row, if one was still going to have a chance to win money on just two in a row.

In any case, we can apply the same kind of analysis to this new example of yours as I did before, and get a premium of $233,000. This is less than the premium of $300,000 doing it your way, because in your example, LA can win three different payouts, for winning two, three and four. I was assuming he could win only one payout, depending on how far in the streak he got. I think I was wrong about this, that the payouts were cumulative?

 

[Benotti69]

Funny thing is these allegedly great legal minds (the best money can rent?) aren't coming up with any brilliant legal strategies commiserate with their billable hours.

Judging from the link Race Radio posted, they are using the type of low-end stalling tactics one would expect from a lawyer like Saul Goodman.

I dont think Armstrong's lawyers are in the legal business for any other reason other than to enrich themselves. They seem to be doing very well at that with Armstrong and the longer it goes on the richer they get.

 

[Bluenote]

I dont think Armstrong's lawyers are in the legal business for any other reason other than to enrich themselves. They seem to be doing very well at that with Armstrong and the longer it goes on the richer they get.

How did no one on his team see the Reasoned Decision going public? Or his ban extending to Triathlons?

Head scratching, really.

 

[Race Radio]

After all his recent legal setbacks it is understandable that lance celebrates the Feds recent mistake

 

[frenchfry]

After all his recent legal setbacks it is understandable that lance celebrates the Feds recent mistake


*deleted*

Was that taken at one of his kid's birthday parties?