The 2012 CQ Ranking Manager Game

[Tank Engine]

I was joking

Thomas the Tank Engine only does deadly serious

 

[The Hitch]

TT:

250
150
120

RR:

400
240
200

The Olympics were somewhat of a strategy in choosing my team. Mainly for the TT.

Assuming that Martin Cancellara Wiggins Contador Evans Porte share the top spots thats not really many points to be basing a pick around.

Excepting Larsson of course. I ended up dropping him in favour of other riders at that price, though now im back to considering him a better pick than Madrazo and am regretting not choosing him

 

[hrotha]

It's not like Cancellara will double his score from last year or anything. If he matches his best ever point total, which is unlikely as he's past his prime, it's still only a profit of 30 percent.

That's a bold statement.

 

[will10]

1000+ point gambles

29 riders had a 2011 value of 1000+, and I'm surprised to see how many have been picked - how much value do you think there is in some of these big point picks?

RIDER 2011 SCORE POPULARITY
GILBERT Philippe 3168 0
CONTADOR VELASCO Alberto 2373 1
RODRIGUEZ OLIVER Joaquin 2028 1
EVANS Cadel 1936 2
VOECKLER Thomas 1744 0
MARTIN Tony 1694 0
SAGAN Peter 1596 3
SANCHEZ GONZALEZ Samuel 1576 0
CAVENDISH Mark 1469 3
SCARPONI Michele 1437 0
CANCELLARA Fabian 1432 6
BOASSON HAGEN Edvald 1391 4
WIGGINS Bradley 1322 1
ROJAS GIL Jose Joaquin 1304 0
SCHLECK Fränk 1292 0
VISCONTI Giovanni 1281 0
VAN AVERMAET Greg 1199 0
GOSS Matthew 1189 2
CUNEGO Damiano 1175 1
GREIPEL Andre 1173 2
SCHLECK Andy 1152 3
NIBALI Vincenzo 1127 9
MOLLEMA Bauke 1101 2
LEIPHEIMER Levi 1074 1
MARCATO Marco 1055 0
COBO ACEBO Juan Jose 1048 0
KITTEL Marcel 1041 1
MARTIN Daniel 1035 4
LEUKEMANS Björn 1019 0

 

[Netserk]

1000+ point gambles

29 riders had a 2011 value of 1000+, and i'm surprised to see how many have been picked - how much value do you think there is in some of these big point picks?

Rider 2011 score popularity
nibali vincenzo 1127 9

Let me see, perhaps because he in 2010 scored 2204 points

 

[sublimit]

I considered Larsson, Rasmussen for the TT points but in the end left them out because the big points will be likeley to be picked up by Martin, Cance, maybe Wiggins, Porte and Froome possibly.

Another tester under consideration was Zabriskie because he was like 160 points or so, and theres plenty of points at the US nationals ITT but in the end he couldn't fit into my team for one reason or another.

 

[LaFlorecita]

1000+ point gambles

29 riders had a 2011 value of 1000+, and I'm surprised to see how many have been picked - how much value do you think there is in some of these big point picks?

<snip>

I'm particulary (still) surprised by Michiel picking Contador. If he's improving it will be just to maybe 2500 points instead of 2300, if he really has an awesome season a bit higher, but there's also still a fair chance he isn't even going to ride this season.

Also, I can't understand why there are some who have chosen Evans. Do they expect him to win both the Worlds and the Olympics? (apart from le Tour)

 

[roundabout]

Cavendish/Wiggins/Goss/Nibali/Leipheimer/Martin seem the most reasonable to me

 

[LaFlorecita]

Another tester under consideration was Zabriskie because he was like 160 points or so, and theres plenty of points at the US nationals ITT but in the end he couldn't fit into my team for one reason or another.

Yeah I considered him as well but in the end I figured he wasn't worth it. (lol, he's not good enough for a spot on my awesome superteam.)

 

[will10]

Cavendish/Wiggins/Goss/Nibali/Leipheimer/Martin seem the most reasonable to me

Where is Wiggins going to get a points increase from??

 

[LaFlorecita]

Where is Wiggins going to get a points increase from??

Tour + Olympics. (at least that's what I suspect. I don't believe in it though)

 

[the asian]

I'm particulary (still) surprised by Michiel picking Contador. If he's improving it will be just to maybe 2500 points instead of 2300, if he really has an awesome season a bit higher, but there's also still a fair chance he isn't even going to ride this season.

Also, I can't understand why there are some who have chosen Evans. Do they expect him to win both the Worlds and the Olympics? (apart from le Tour)

Must have been selected by devoted fans just for fun.

Not all participants participate to win.

 

[roundabout]

Where is Wiggins going to get a points increase from??

Olympics. Tour.

Maybe the Sky TTT (not sure how it will be scored and if he rides it)

On the other hand there won't be such a TT friendly Paris-Nice this year even though I believe that he could do well even with a climb in Mende and he may not win the Dauphine again.

 

[the asian]

Cavendish/Wiggins/Goss/Nibali/Leipheimer/Martin seem the most reasonable to me

Where is Cav going to pick up more points??

Olympic RR win will be cancelled out by Worlds RR.

Are you expecting him to win MSR, Roubaix?

 

[mc_mountain]

1000+ point gambles

29 riders had a 2011 value of 1000+, and I'm surprised to see how many have been picked - how much value do you think there is in some of these big point picks?

the value of these big riders is twofold (and applies to Cancellara for this year)...

1. uses up points, allowing for the selection of more bargain riders, with the potential to triple or better their score. they dont need to be a huge profit to be useful (last year I had EBH and Cav with a combined profit of c.400 and came 3rd in the game)

2. high profile wins/results which is good for the sponsor and team morale (ie enjoyment in the game)

As it happens I've not repeated this policy this year (if poels and pinot have great years and renshaw and pozzato dont score better than say 500 each, I think my team can do really well).

having said that I'm surprised that some people have gone for Basso, as there doesnt seem to be much chance of providing a profit of more than say 200 (I'd hoped for more from Cav and EBH). 2009 - 2x good GTs 1100, 2010 - 1x excellent GT 1000, 2011 - 1x good GT 900 (appreciate this is a simplification of his scoring). The risk/reward doesnt seem to be worthwhile to me, whereas someone like nibali or farrar would could score 500-1000 points more I think are good picks.

I dont really rate the prospects of Renshaw and Pozzato scoring really big (but understand the arguments that they might), and if they dont do well that puts me in a really good position, seeing how popular they have been though will make for a nervous season.

if you look at Cav's stated goals for the season - MSR, Gent Wevelgem, Giro Points jersey, green at the tour, olympics, he could easily score 2000 if can hit these albeit lofty (and some unlikely) ambitions.

 

[roundabout]

Where is Cav going to pick up more points??

Olympic RR win will be cancelled out by Worlds RR.

Are you expecting him to win MSR, Roubaix?

There are more flat stages in the Giro and while I have some doubts that he can actually win Sanremo I think he can score good points there.

http://cqranking.com/men/asp/gen/rider_palm.asp?riderid=3448&year=2011&all=1&current=0

I think he can score more points in the spring at least.

 

[LaFlorecita]

There are more flat stages in the Giro and while I have some doubts that he can actually win Sanremo I think he can score good points there.

http://cqranking.com/men/asp/gen/rider_palm.asp?riderid=3448&year=2011&all=1&current=0

I think he can score more points in the spring at least.

You've picked Cav, I suppose?

 

[18-Valve. (pithy)]

That's a bold statement.

Maybe. Canc is past his prime when it comes to TTs, though. If Martin doesn't get caught or injured it's over.

 

[roundabout]

No. I compete for fun.

 

[Hugo Koblet]

mc_mountain is right when he says that some of the value comes from using up points to allow more bargains.

For instance, the best possible team last year would have included Gilbert even though he "only" managed a 170% profit and the overall profit of the best team would be something like 350% or so.

The best 1000+ pointers I think are Goss, Nibali, D. Martin in that order.

 

[Ferminal]

1000+ point gambles

29 riders had a 2011 value of 1000+, and I'm surprised to see how many have been picked - how much value do you think there is in some of these big point picks?

RIDER 2011 SCORE POPULARITY
GILBERT Philippe 3168 0
CONTADOR VELASCO Alberto 2373 1
RODRIGUEZ OLIVER Joaquin 2028 1
EVANS Cadel 1936 2
VOECKLER Thomas 1744 0
MARTIN Tony 1694 0
SAGAN Peter 1596 3
SANCHEZ GONZALEZ Samuel 1576 0
CAVENDISH Mark 1469 3
SCARPONI Michele 1437 0
CANCELLARA Fabian 1432 6
BOASSON HAGEN Edvald 1391 4
WIGGINS Bradley 1322 1
ROJAS GIL Jose Joaquin 1304 0
SCHLECK Fränk 1292 0
VISCONTI Giovanni 1281 0
VAN AVERMAET Greg 1199 0
GOSS Matthew 1189 2
CUNEGO Damiano 1175 1
GREIPEL Andre 1173 2
SCHLECK Andy 1152 3
NIBALI Vincenzo 1127 9
MOLLEMA Bauke 1101 2
LEIPHEIMER Levi 1074 1
MARCATO Marco 1055 0
COBO ACEBO Juan Jose 1048 0
KITTEL Marcel 1041 1
MARTIN Daniel 1035 4
LEUKEMANS Björn 1019 0

My top 3 picks are fanboy picks and because I want to have some big names in my team. Cancellara will do OK, but he will not be in the best team. It also happened to be the case that I had filled a team with riders who I think will go well, and had 2500 points remaining.

 

[The Hitch]

Let me see, perhaps because he in 2010 scored 2204 points

Nibali can definitely improve on his current score especially with good performances in MSR Liege Lombardia and of course worlds.

Tje reason I never have serious thought to him, other than my no riders over 620 policy, was that in 2010 he did Giro Vuelt whereas next year he is doing a tour that doesnt suit him (however much cineteq tries to claim Nibali is the best gc tter).

 

[mc_mountain]

mc_mountain is right when he says that some of the value comes from using up points to allow more bargains.

For instance, the best possible team last year would have included Gilbert even though he "only" managed a 170% profit and the overall profit of the best team would be something like 350% or so.

The best 1000+ pointers I think are Goss, Nibali, D. Martin in that order.

good stat Hugo, that illustrates the point really well. on the flipside, it would be very difficult to find 33 riders at circa 222 points that will all triple their scores.

 

[Skibby the bush kangaroo]

I meant something like top 5 if the history repeats. As mentioned above, last years most popular picks would give you 3rd place.

Same for Skibby, his calculation for his 2012 team is probably correct and the software picked the team based on some basic conditions but many other important (risk factor) conditions were not applied otherwise he would not have Nibali or any other expensive rider on his team....unless I'm wrong and he did pick his team manually

Yes you're right, I basically fitted a model to all past data (2002-2011 seasons) using CQ points in previous 3 seasons, age, team status (Pro-tour, Pro-conti, Cont and Elite-3) and change in status (e.g. up to PT or down from PT) as covariates. Got predictions on 2012 points based on these and then optimized. So it pretty much is looking at people who dropped points in 2011 compared to 2010 but is also accounting for 2009 and also weighting younger riders with drops more highly (or letting them have smaller drops).

As you say, Nibali might not be the best choice if more information could be factored into the model. I think the model could be improved quite a bit as it pretty much disregards anyone who has a 0 or close to 0 season as being too much of a liability.

Incidentally, even if my model were completely correct (which it obviously isn't) my estimated probability of winning is only 11% (taking into account everyone's choices and my model estimates for those riders). I picked a team based on this method for 2011 and it would have come top 15 but not top 10 (don't have the precise team to hand now can find it tonight), but the gained points were within the prediction error bounds from the model.

 

[Skibby the bush kangaroo]

Skibby, that's some excellent work indeed! Thanks! I'm working on some sort of "article" so I'll probably use some of your statistics if that's OK!

Yeah no problem. Let me know if you want any thing else calculated as well.