The 2023 CQ Ranking Manager Thread

[ingsve]

- Ingsve would be 29th by that measure with only Kristoff, Viviani and Uran still going but that doesn't take anything away from their win since nobody is trying to optimize their team for 12 years down the road.

Ya, that probably tracks with my strategy. Back then I was more looking for proven riders with a track record but who had a bad year rather than finding those really young riders who are more likely to still be around in 2023.

 

[ingsve]

If expected value is the same, I think it's rational to be risk seeking in this game. So the more eggs you can fit in one bag, the better.

Ya, I suppose it depends on what the overall goal with the strategy is. If you are going for the win and nothing else then yes, putting more eggs in one basket is likely a good strategy. If the goal is to have a team that will more consistently be part of the top then spreading the risks is probably better.

 

[ingsve]

But this is what I was trying to factor in by scaling the optimal team ratio down to 'if the 7500 team is good enough to double, what is the % return of an equally strong 5672 team' and then seeing what the expensive rider had to make up to reach 15000. By scaling both teams' % return down based on the ratio between the respective optimal teams for both budgets, I factored in the assumption that none of those teams are likely to find all the best picks. There are mid-level picks who also increase massively and who could be missed, so at least in my mind, this calculations holds water. But I'd be interested in hearing a different take. It's a bit hard to understand exactly what you mean here though.

I guess the way I think of it is whether there is a difference in how easy it is to find the good picks in the midrange cost or the low cost region. The ratio you used based on the optimal composition means there is an equal chance at finding the good picks because if there was a good pick then the optimal comp had them. In reality my gut feeling at least would be that it's easier to find good picks in the midrange than in the low cost range simply because in the midrange cost you have more information to go by since the rider has already performed more on the relevant stage whereas at the low cost level there are so many different riders to choose from and less information to go by that it makes it harder to pick right. That would skew the ratio you used I believe which would lower the overall output of the 5672 part of the team meaning the star would have to compensate more.

I could ofc be wrong in that assumption. I don't really have data to back it up, just my gut feeling.

And if the 'expected score' for both sets is equal, then for set 2 to be a success, you can't tolerate any of the picks to really bomb. But for set 1, you mainly need the one expensive rider to perform as you expect.

Is it really right to view the scenarios as having equal expected value though? With an expensive rider there is often a cap on how much that rider can potentially return. The midrange options will always have a higher ceiling to grow under whereas the expensive rider will never be able to tripple their score most likely. That would skew the expected values wouldn't it?

 

[Squire]

Is it really right to view the scenarios as having equal expected value though? With an expensive rider there is often a cap on how much that rider can potentially return. The midrange options will always have a higher ceiling to grow under whereas the expensive rider will never be able to tripple their score most likely. That would skew the expected values wouldn't it?

Of course if you think there's a higher ceiling for the mid-range picks and you think that betting on them collectively achieving it is worthwhile compared to the one expensive rider, then the expected values aren't the same. My example was made on the premise that expected values were exactly the same, just so I could illustrate that there is just as much (if not more) risk in spreading the eggs across more baskets. In a real scenario there might not be an expensive pick available with enough upside, or there might not be enough good mid-range ones to make that strategy worth it. I was just questioning the assumption that a very expensive rider = more risk by default.

Your gut feeling about how easy/hard it is to find picks in different price ranges is interesting. But for every cheap pick that makes the optimal team, there are also quite a lot of other cheap picks with slightly less but still massive percentage increase. So I don't know the answer to that one, really.

I've just noticed that for my own teams, it's often the 300-500 point riders I pick that are the most frustratingly bad picks. There's a much smaller pool of riders to chose from in that range, and you might even have years where very few of those do well enough, so in my mind it's gonna be hard to pick the correct ones. Whereas for cheap riders, yes in theory there is a huge number of different riders you can pick, but in reality you can quite easily narrow it down to a manageable list of those who are somewhat likely to do well.

Again, not really sure if any direction is more correct than the other here, but at least I don't think my ratios based on the optimal teams are that far off.

 

[EvansIsTheBest]

A couple of years ago there was a discussion about this, and I made some calculations that showed that an expensive rider hardly had to make a profit for you to win the game, provided that the rest of your team was of a 'normal' game-winning quality. You can read it here.

I wonder how much the math changes on expensive picks if you move the target score around. Because 15k might win you the game sometimes but most years it's "just" a good team. To be fair, it's a very good middle ground (throwing out the pandemic affected 2020 and 2021 years, the average and median winning score are both around 15800 points) and you have no real way to know beforehand if it's going to be a high or low scoring year. But I would imagine knowing in advance if you need 13k or 18k to win a particular year would greatly impact whether an expensive pick makes sense.

 

[EvansIsTheBest]

Circling back to what I was discussing earlier with squire about whether you should expect winning teams to continue outperforming their competition in following seasons. I did a bit of number crunching and here’s the result:

Team​	Year​	Year N+1​	Year N+2​	Year N+3​	Year N+4​	Year N+5​
ingsve​	2011​	5th​	34th​	17th​	21st​	35th​
skidmark​	2012​	1st​	5th​	9th​	13th​	12th​
SteelyDan​	2013​	39th​	35th​	11th​	41st​	29th​
Pentacycle​	2014​	5th​	24th​	53rd​	57th​	55th​
Cykeltyven​	2015​	5th​	15th​	23rd​	27th​	32nd​
skidmark​	2016​	9th​	13th​	28th​	18th​	19th​
fauniera​	2017​	5th​	10th​	11th​	3rd​	7th​
ruvu75​	2018​	25th​	23rd​	11th​	12th​	25th​
fauniera​	2019​	26th​	23rd​	51st​	29th​	TBD​
bminchow​	2020​	3rd​	24th​	21st​	TBD​	TBD​
Object​	2021​	1st​	2nd​	TBD​	TBD​	TBD​
EvansIsTheBest​	2022​	1st​	TBD​	TBD​	TBD​	TBD​
Blues in the bottle​	2023​	TBD​	TBD​	TBD​	TBD​	TBD​

Scattered thoughts:
- The winning teams tend to only stay good for an extra season after their win. Presumably they are then overtaken by teams that went with more of a youth oriented setup.
- The three teams that failed to crack the top 10 in the season following their wins had 3 of the 4 lowest winning margins ever which explains why they fell back to the middle of the pack so easily.
- Skidmark 2012, Object 2021 and EITB 2022 are the only teams to “repeat” their wins in the following year. They were also the three highest winning scores ever. Better teams stay good longer.
- In 2021, everybody stacked their teams with veterans that were way underpriced due to Covid so it’s less likely that Object will drop behind youth focused teams in the future. It should become, in time, the most consistent winning team in this metric.
- To be clear this is where they would have ranked if people had kept re-submitting the same team year after year with no regards to budget. It's not where they would have ranked versus the teams that were actually submitted in those years.

 

[Squire]

An interesting thing is that SteelyDan's team won in a year with a low-scoring optimal team, and kinda 'came good' again three years later when there was another year with a low-scoring optimal team. I guess that could indicate that 2013 and 2016 were years with not too many big breakthroughs from young riders.

 

[Armchair Cyclist]

It is not too late to enter the Cyclocross manager Game. It's not a CQ points game (because CQ doesn't do cyclocross points), nor does it have the same type of budget: just give 18 names (men or women), with a restriction on how many you can select from the top 6, and the top 12, of last season's rankings.

You have already missed some points from the Koppenberg Cross, but if you get your team in to me before the Women's Elite race at the Euros starts, then you will be able to get points for that. All details in the link on the top line of this post. And it really will be much quicker than the other games to get a team together.

 

[RedheadDane]

- To be clear this is where they would have ranked if people had kept re-submitting the same team year after year with no regards to budget.

And then of course you'd eventually run into another problem.
Let's say someone picked Nibali, Gilbert, Valverde, and Kittel for the 2013 game. They would have been quite useless this year.

 

[ingsve]

I've just noticed that for my own teams, it's often the 300-500 point riders I pick that are the most frustratingly bad picks. There's a much smaller pool of riders to chose from in that range, and you might even have years where very few of those do well enough, so in my mind it's gonna be hard to pick the correct ones.

Ya, I suppose you have the effect of riders reaching their level in the midrange. They seem to be on an upward trajectory for a couple years and you think they might continue and reach 1000+ level or higher but they end up capping out at 500 instead of increasing.

 

[Anderis]

They seem to be on an upward trajectory for a couple years and you think they might continue and reach 1000+ level or higher but they end up capping out at 500 instead of increasing.

They may be a Piccolo. You take them because they only raced for half a season the year before, so even if they don't improve their level, they should be able to score almost twice the points by the sole reason of riding a full calendar. You end up finding that they forgot how to finish races anywhere near the top places and they don't even reach 20% return.

 

[EvansIsTheBest]

Ya, I suppose you have the effect of riders reaching their level in the midrange. They seem to be on an upward trajectory for a couple years and you think they might continue and reach 1000+ level or higher but they end up capping out at 500 instead of increasing.

That could be it but I'm not necessarily sure that riders in that range turn out to be bad/average picks more often than others. There might be some bias in that it's easy to remember the expensive guys that disappoint but when a cheap pick underperforms it's forgotten pretty much instantly.

 

[Armchair Cyclist]

Last call for the 2023-24 Cyclocross Game. Submit a team before Saturday's Jaarmarktcross WE race to get points for that event, or at least before Sunday's Dendermonde start to meet the final deadline.

 

[hfgon1]

Can we do a 2024 CQ Ranking Manager Game with a team in which to select riders with the highest point losses? Think that might be my online chance to win it once.

Well... As I think about it now... Should be an interesting game. Just let riders above 32yrs of age, retiring or doped ones out and it would be hard to select. Sepp Kuss would surely be popular

 

[DJ Sprtsch]

Can we do a 2024 CQ Ranking Manager Game with a team in which to select riders with the highest point losses? Think that might be my online chance to win it once.

Well... As I think about it now... Should be an interesting game. Just let riders above 32yrs of age, retiring or doped ones out and it would be hard to select. Sepp Kuss would surely be popular

There was a game like that some years ago. The career in decline game. (can't beleive it's been that long ago!).
I found it highly entertaining, somehow it very much appealed to me.

 

[Armchair Cyclist]

Yeah, the problem was tracking that riders had done enough races to qualify as declining rather than effectively retired.

Reviving it is not on my to do list, but if anyone really wants to and I can help with unpacking the spreadsheet, I'll look at it again.

 

[skidmark]

Hey everyone! Hope you're hard at work on your 2024 teams. Just a note to say:

1) New thread should be up about the middle of December, as per usual, so keep your eyes peeled;

2) Regular doping violation rules will apply to all riders, no exceptions made for specific cases. Thanks to everyone who weighed in on the specific case discussed here; I agree with the assessment of the majority, even if it will mean that one rider is likely to be picked quite a bit.

Just wanted to make sure the rules are clear on that!

 

[EvansIsTheBest]

Hey everyone! Hope you're hard at work on your 2024 teams. Just a note to say:

1) New thread should be up about the middle of December, as per usual, so keep your eyes peeled;

2) Regular doping violation rules will apply to all riders, no exceptions made for specific cases. Thanks to everyone who weighed in on the specific case discussed here; I agree with the assessment of the majority, even if it will mean that one rider is likely to be picked quite a bit.

Just wanted to make sure the rules are clear on that!

There's still something I need some clarification on. What are two good, obscure riders for a combined price of 60ish points ? Asking for a friend of course.

 

[Devil's Elbow]

There's still something I need some clarification on. What are two good, obscure riders for a combined price of 60ish points ? Asking for a friend of course.

My sources tell me Chris Froome is still available for a really cheap price. Don’t share with anybody else, though.

 

[Salvarani]

Most of the team is done for next year... just the usual start over in final days left.

 

[ingsve]

My sources tell me Chris Froome is still available for a really cheap price. Don’t share with anybody else, though.

I'm going with Cobo. He was good for me in 2011 and this year his price is also 0 I think.

 

[pman]

I want to join this for next year (I think my wife who doesn't post on the forum even wants to do this too). I've been following this a bit so I mostly understand how the game works. I am just confused- do you keep the same team year after year? Can new players join? Thanks!

 

[EvansIsTheBest]

I want to join this for next year (I think my wife who doesn't post on the forum even wants to do this too). I've been following this a bit so I mostly understand how the game works. I am just confused- do you keep the same team year after year? Can new players join? Thanks!

It's always nice to see new faces around here so new players are welcomed. You submit a different team at the beginning of every season. You can pick a rider multiple years in a row but you don't have to. The only requirements are staying under budget (7500) and picking the right number of riders (33 or less) as laid out in the first post of the thread. Here are my last two teams so you can get an idea of what that looks like:

Spoiler: 2023 team

ROGLIC Primoz​

AYUSO PESQUERA Juan​

PIDCOCK Thomas​

PICCOLO Andrea​

ALAPHILIPPE Julian​

JOHANNESSEN Tobias Halland​

LAURANCE Axel​

ASGREEN Kasper​

TESSON Jason​

UIJTDEBROEKS Cian​

PAGE Hugo​

SCHACHMANN Maximilian​

MIHKELS Madis​

VACEK Mathias​

VAN UDEN Casper​

VAN EETVELT Lennert​

PENHOËT Paul​

MARTINEZ Lenny​

ARRIETA LIZARRAGA Igor​

GREGOIRE Romain​

NYS Thibau​

POOLE Max​

SYRITSA Gleb​

GLOAG Thomas​

MOSCON Gianni​

VERCHER Mattéo​

HALVORSEN Kristoffer​

ROMO OLIVER Javier​

HERMANS Ben​

CONTI Valerio​

BERNAL GOMEZ Egan Arley​

DEKKER David​

TARLING Joshua​

Number of riders: 33
Cost: 7498

Spoiler: 2022 team

EVENEPOEL Remco​

EWAN Caleb​

GIRMAY HAILU Biniam​

HIGUITA GARCIA Sergio Andres​

SKJELMOSE JENSEN Mattias​

BAGIOLI Andrea​

RODRIGUEZ CANO Carlos​

VERMEERSCH Florian​

MARTINEZ POVEDA Daniel Felipe​

DUMOULIN Tom​

BATTISTELLA Samuele​

SIMMONS Quinn​

KOOIJ Olav​

PINOT Thibaut​

TULETT Ben​

BUITRAGO SANCHEZ Santiago​

BARONCINI Filippo​

GONZALEZ RIVERA Abner​

AYUSO PESQUERA Juan​

ZINGLE Axel​

VAN WILDER Ilan​

HINDLEY Jai​

VAN GILS Maxim​

VAN TRICHT Stan​

QUINN Sean​

GROVES Kaden​

KÄMNA Lennard​

MÜHLBERGER Gregor​

TOUMIRE Hugo​

BRENNER Marco​

UIJTDEBROEKS Cian​

RIVERA SERRANO Kevin​

KORETZKY Victor​

Number of riders: 33
Cost: 7498

 

[pman]

Thanks!

 

[Squire]

Most of the team is done for next year... just the usual start over in final days left.

Yeah, same. Most of my team has been quite settled for at least a month now, but the final 2-4 spots are a major headache. There's a price range where I really would have wanted another good pick when there are none, to make everything fit. I either have a team I'm really happy with that's a few hundred below budget (obviously not feasible) or a slightly unsatisfying team near budget.

I'm trying to tell myself to let it be until more information is available closer to the deadline, but it's hard not to come back to it from time to time!