Velominati Super Prestige

View Comments

• Scratch that. It's coming back together now.

• Frank... where do you want the graphs ?

  I've done some analysis of the main races - not all. Paris-Nice, Ronde, Roubaix, Liege and Giro with TdS to come.

  I concentrated on the VSP picks compared to race results rather than the actual VSP competition - honestly I don't know how anybody scores any points after looking at it. Two objectives:

  1. Whether being in the Velominati VSP or Top 10 is an indicator of actual results.

  2. Is there any correlation between the strength of picks and results.

  I have defined correct results as simply being Top 5 (VSP) or Top 10, not exact placing.

   

  Method

  I calculated the weighted votes, so 5 points for each first place pick, four for second and so on to produce a Velominati VSP and a Velominati Top 10.

  Weighted votes were also calculated for riders finishing in the Top 10 of actual placings, where they were not already in the Velominati selection.

  I also calculated the percentage of votes for each rider in the VSP and Top 10, as a share of the total, to indicate and compare the strength of preference e.g. for Paris Nice the VSP total weighted votes was 1183, of which Wiggins had 378 or 32%. Whereas for Liege the total weighted votes was 1220, of which Gilbert had 527, or 43.2%.

   

  Data

  I won't put all the numbers here but for interest the collective Velominati VSPs were:

  Paris Nice: Wiggins (32%), Valverde (24.3%), T Martin (22.5%), Leipheimer (11.3%), F Schleck (9.9%)

  Ronde: Boonen (34.4%), Cancellara (33.3%), Sagan (16.5%), Chavanel (10%), Gilbert (5.8)

  Roubaix: Boonen (38.3%), Pozzato (23%), Ballan (15.8%), Hushovd (15.6%), Flecha (7.3%)

  Liege: Gilbert (43.2%), J-Rod (26.6%), F Schleck (10.4%), Valverde (10.1%), Cunego (9.8%)

  Giro: Scarponi (32.2%), Basso (25.7%), Kreuziger (21.1%), J-Rod (12.1%), F Schleck (8.9%)

  TdS: Gesink (30.5%), Cunego (21.6%), Kreuziger (18.2%), Valverde (15.2%), Leipheimer (14.5%)

   

  Analysis

  Comparing the VSPs and Top10s to results was the basis of the analysis - the number of VSP picks actually in the Top 5, and likewise the number of Top 10 V picks in the actual Top 10.

  Actual placings in the Top 10 but not within the Velominati Top 10 were also analysed.

  This was done in two ways. First by adding the total percentages for the correct picks, and secondly by looking at the outright number correctly picked.

  This would show as a higher weight-based score if two very strong favourites were picked in the Velominati VSP and both placed Top 5, compared to a more evenly spread Velominati VSP selection but which resulted in the same number of more riders placing in the Top 5. The intention was not only to show how many VSP picks (and Top 10) were correct, but the extent to which they had been correctly picked (or conversely, the extent to which the VSP picks were wrong).

  For example Gilbert and Rodriguez between them attracted high weighted VSP vote for LBL and had only those two (not Schleck, Valverde and Cunego) been in the Top 5 results the score would have been 69.8% and an outright score of 2 (from 5). Whereas had Schleck, Valverde and Cunego placed in the Top 5 the outright score would be higher (3 from 5) but the weighted result would be only 30.2%, indicating a weaker favouritism and comparatively poor result.

  Or as an example from the other direction, the actual number of Velominati Top 10 picks who gained placing in the Top 10 at both Roubaix and Liege was 4, but the percentage allocated to them at Roubaix was 54.2% and Liege 13.4%, reflecting a weaker performance.

   

  Results

  This is where I need a table but the results were:

  Weighted % VSP: P-N 56.3%; Ronde 61%; Roubaix 61.4%; Liege 0%; Giro 70%. Average 49.8%.

  Weighted  % Top 10: P-N 49.9%; Ronde 56.5%; Roubaix 54.2%; Liege 13.4%; Giro 69%. Average 48.6%

  VSP Correct Picks: P-N 2; Ronde 2; Roubaix 3; Liege 0; Giro 3. Average 2.

  Top 10 Correct Picks: P-N 3; Ronde 5; Roubaix 4; Liege 4; Giro 5. Average 4.2.

  Finally I looked at the number of riders who actually placed in the Top 10, but were not in the Velominati Top 10 to see the amount of weighted picks they scored, in case it indicated a significant but not large amount of favouritism. it didn't.

  Non V10: P-N 1.5%; Ronde 2.7%; Roubaix .5%; Liege 4.8%; Giro 2.4%. Average 2.4%.

   

  Conclusions

  The Velominati collectively pick just two of the Top 5 in each competition to date, and around four of the top 10. Just under half the weighted votes cast are 'correct' in the sense that they go to a rider who places accordingly, if not exactly.

  The Giro was our best performance to date with 3 VSP and 5 Top 10, and around 70% of votes correct in both.

  Liege was an aberration with not a single VSP rider placing in the Top 5.

  In general you would have a better chance by avoiding riders who do not feature in the VSP top 5 or 10, although the problem is then which of the other riders to choose.

   

  Future Research

  Having observed this we may now see an illustration of the Hawthorne Effect (or Observer Effect if the scientists would prefer) by causing a modification in behaviour in future VSPs.

  it would be of interest to look at the behaviour of VSP participants in terms of the timeline of votes to determine whether there is a suggestive process caused by the open nature of the selection.

  In regard to results the votes of individual riders could be looked at over the events and the selection in individual ranks. For example riders like Frank Schleck and Damiano Cunego are consistently picked in or near the Top 5 by weight of votes but rarely in the top one or two placings.

  If anyone can award me a PhD or even a Masters I'd be happy to do more...

• @KeepersWhen is Le Tour VSP opening?The only reason I ask is that I'm off on holiday next day and won't have internet access without looking for it. Alternatively is there anyway I can submit my picks now? Cheers

• Starting to look at 2012 Tour de France start list here.

• Here's to a set of leggers about to hit the Tour. No mention of Johan Bruyneel.

• @versio JB is not going to be in attendance. He's sitting this one out in light of his alleged connection to the ongoing investigation by the USADA into Pharmy. He doesn't run the team from the car in the road anyway, that's done by sports directors Dirk Demol and Alain Gallopin.

  http://velonews.competitor.com/2012/06/news/bruyneel-to-sit-out-tour-de-france-over-usada-allegations_225383

• @scaler911 Yes, of course. Supposed to hear the lampooning in "No mention of Johan Bruyneel."

• I hope we're not forgetting the Giro d'Italia Donne. I've just realised that the first stage is going to pass nearby my house. Definately won't be one of the more scenic moments of the cycling calendar. They'll be glad of the police escort as they'll be passing by a housing estate know locally as 'o terza mondo' 'the third world.'

• 350 entrants in the VSP!

• O I am on hols from 28th any chance the Olympic Road Race picks can be put up a little early ;)