Decathlon-style multi-event score from logbook

[Elizabeth]

Interesting stuff!

Just to be clear: the sample file above is for the past season (2022-2023).

Ah ok. Was wondering, because I could not find my entry by scrolling through. But now I scrolled more way down to find position 378 and see that there is no 2k time, but I definitely had ranked this piece with a 6:45.

I'm 463, missing a 5k time. It looks like my 5k was "race" and not "verified". There are some also races that don't seem to push out to the logbook, and I have seen some very strong unverified pieces that I know were done in competition. Dena will help with verifying if people reach out, but not everyone cares.

C2 only allows you to rank one piece per distance, so it's not possible to rank both a slides piece and a static piece, and I initially assumed that was the issue. For distances 2k+, I don't personally think that slides offer an advantage. I'm not sure if this is the commonly accepted thought. They absolutely offer an advantage at sprints.

[HornetMaX]

I'm 463, missing a 5k time.

That is, across any gender I guess

It looks like my 5k was "race" and not "verified". There are some also races that don't seem to push out to the logbook, and I have seen some very strong unverified pieces that I know were done in competition. Dena will help with verifying if people reach out, but not everyone cares.

Yeah, races are a bit of a problem as the entries do not have the UserID, so I cannot properly pivot them ("merge" them with non-race entries of the same athlete). I could try to merge on athlete name only, but even this is not 100% granted to work (slightly different names, fancy chars etc).

C2 only allows you to rank one piece per distance, so it's not possible to rank both a slides piece and a static piece, and I initially assumed that was the issue. For distances 2k+, I don't personally think that slides offer an advantage. I'm not sure if this is the commonly accepted thought. They absolutely offer an advantage at sprints.

That's why I can chose what to consider when creating the final ranking: it's simply a command line option when creating the final rankings file.
I'm open to any advice from savvy rowers on what makes sense.
First thing would be if same scoring for M and F is OK: I like it as it allows to better compare M & F rowers, but that's just my 2c.
If consensus is to score F according to F WRs, I can change that easily.

[Elizabeth]

Yeah, races are a bit of a problem as the entries do not have the UserID, so I cannot properly pivot them ("merge" them with non-race entries of the same athlete). I could try to merge on athlete name only, but even this is not 100% granted to work (slightly different names, fancy chars etc).

Oh! I didn't realize that, but now understand why we will see (say) Viv Culverwell and Vivian Culverwell. I think Libby Garbis's race results have been under her given name Elizabeth, but see that she does have ranked distances at each of the ones you're looking at.

I don't feel strongly about having male and female times mixed in, but think a final result that lets you filter by age/sex/etc would be helpful. Are you using the open world records, or a certain age group amongst the men?

If this is more of a general-interest tool, it would make sense to me to have it look at a couple of additional "similar" distances and then choose the best to rank. This could help fill some of the gaps. What I mean is that some teams choose a 5k and some choose a 6k for a season time trial, and performance at one would roughly correlate with performance at the other. So either could "count". If it's something to provide some motivation to get in the given ranked distances and see how they stack up, then it makes more sense to leave those numbers at zero.

I think the biggest distances that I see people focus on are 500m, 2k, 5k/6k, and 30r20. I can't remember when I last did a 10k. But 30r20 are not rankable as such. (I currently have one ranked as a 30:00 piece though.)

[HornetMaX]

I don't feel strongly about having male and female times mixed in, but think a final result that lets you filter by age/sex/etc would be helpful. Are you using the open world records, or a certain age group amongst the men?

That's definitely the case, I can filter by age/sex/etc in the final result, so that the #1 W is #1 when filtering by gender and maybe #257 when not filtering by gender (i.e. there are 256 M rowers faster than her).

The question is more: assuming all the overall WRs are hold by M rowers, in a 3-events c2athlon a top-notch M would have a score close to 3000pts, while a top-notch F would have much less (because I use M WRs are reference).

If this is more of a general-interest tool, it would make sense to me to have it look at a couple of additional "similar" distances and then choose the best to rank. This could help fill some of the gaps. What I mean is that some teams choose a 5k and some choose a 6k for a season time trial, and performance at one would roughly correlate with performance at the other. So either could "count". If it's something to provide some motivation to get in the given ranked distances and see how they stack up, then it makes more sense to leave those numbers at zero.

The 5K vs 6K seems to be the only instance of event that are so close that rowers can pick one of the two at random (I use the WR as listed here: https://www.concept2.com/indoor-rowers/ ... ords/world).
Probably doable, but would make things a bit more complex on my side. But yeah, it's just code, it's MEANT to be messy

I think the biggest distances that I see people focus on are 500m, 2k, 5k/6k, and 30r20. I can't remember when I last did a 10k. But 30r20 are not rankable as such. (I currently have one ranked as a 30:00 piece though.)

I have no data for 30r20 WR, so that's a no-go I was also thinking about making this a bit funny, with categories like:

• The Boring Core: 2K + 5K
• The Mighty Spread: 500m (or maybe 1min) + HM (or even FM)
• The ICanDoItAll: 500m + 1K + 2K + 5K + 10K + HM + FM

And that before mixing RowErg event in too

[MPx]

One sample file: 2023 ranked workouts, combined scores for rowerg 2K + 5K + 10K (only rowerg, no slides or dynamic, only verified entries): https://docs.google.com/spreadsheets/d/ ... drive_link

Looked at this and I'm not in it. No loss there, but why? All my log entries are verified and all ranked pieces are simply that, rowerg only, not races or subject to name change. Are the WR comparisons age related or simply the open WRs?

[HornetMaX]

Looked at this and I'm not in it. No loss there, but why? All my log entries are verified and all ranked pieces are simply that, rowerg only, not races or subject to name change. Are the WR comparisons age related or simply the open WRs?

Checking why Salky's 2K is not there I've found a nasty bug, my bad. Fix ongoing.

Bad news: fetching all the pages is going to be a pain time wise ...

[HornetMaX]

OK, some good news: bug fixed and fetching is not too slow (if I parallel fetch the different pages).
Also, I seem to have no more issues with c2 ranking pages duplicating/dropping some entries: if they fixed it, that's great.

Finally, I've seen that some athletes have multiple entries for the same season and same event, with different rower types (rowerg, slides, dynamic) and with different weight classes. So, unless I ask for a ranking with specific rower type (or specific weight classes), I now take the best result for the athlete on the event (if he/she has multiple ones). Also keeping the age (min age, if the athlete has entries with different age in the season) and and the gender (uh, here I expect this not to change during a season ) in the output.

For seasons 2022, 2023 and 2024 (up to today, of course) I've uploaded:

• The grabbed rankings (c2R-2022/3/4.csv)
• The rankings for a c2athlon r2K + r5K + r10K (C2Athlon-2022-r2K+r5K+r10K-VerifiedOnly.csv etc)

in that folder: https://drive.google.com/drive/folders/ ... drive_link

They are self-explanatory .csv files.

[Sakly]

Now on 67th for 2023 with the 2k included
Seems to work better now

[MPx]

OK, some good news: bug fixed ...

Yay, I'm in! Only checked 2023 and no 312 doesn't sound too special but I'm there so happy days.

[HornetMaX]

Before calling it a day I want to check if it's possible/worth to also consider race results (meaning that if you have ranked a 2K and a 5K at home but a 10K only in a race, the 10K counts towards the "2K+5K+10K" c2athlon). It won't be perfect (as it relies on the user name entered in the logbook to be IDENTICAL to the one entered in the race result), but better than leaving our race guys out completely

[HornetMaX]

Before calling it a day I want to check if it's possible/worth to also consider race results (meaning that if you have ranked a 2K and a 5K at home but a 10K only in a race, the 10K counts towards the "2K+5K+10K" c2athlon). It won't be perfect (as it relies on the user name entered in the logbook to be IDENTICAL to the one entered in the race result), but better than leaving our race guys out completely

Sorry for our racers, but I took a look and matching on names is very unreliable: we have 4 Scott Anderson in the 2023 season, 2 of which are in the USA.
So the only sensitive ting to do is to use only entries that do have a UserID.

Contrary to what I said before, I just noticed that *some* race entries do have an UserID: these will be kept, but race entries without UserID will be thrown away.

[HornetMaX]

Sorry, didn't have a lot of time recently to play with this. Anyway, I'm more or less done.

I have uploaded some samples in this Google Drive folder: link (accessible to anybody).

So what's in there:

1. Files C2R-2021.csv ... C2R-2024.csv are the 2021-24 rankings (2024, up to today, the season is still ongoing). I can easily import past seasons (c2 data goes all way back to 2002 and 2010 for Rower and SkiErg) and periodically refresh the data of the ongoing season.
2. For each season there are 4 c2athlons:
   • TheCore: included events at 2k, 5k and 10k.
   • TheSmallSplit: included events are 1k and HM.
   • TheBigSplit: included events are 500m and FM.
   • TheFullRange: included events are 500m, 1k, 2k, 5k 10k HM and FM.
3. All files are plain .csv, you can open them with any text editor or spreadsheet software (MS Excel, Google Drive, whatever can import a csv).
4. For each c2athlon there are 7 files:
   • 4 are the combos [Male, Female] x [HW, LW].
   • one each for [Male] (no matter the weight), [Female] (no matter the weight) and [all together] (no matter gender nor weight).

The file naming should be self-explicit, examples:

• "C2R-S2023.csv.TheBigSplit-R-M-L.csv" means 2023 season, BigSplit c2athlon, -R for Rower, -M for Male, -L for Lightweight.
• "C2R-S2024.csv.TheFullRange-R.csv" means 2024 season, FullRange c2athlon, -R for rower (any gender, any weight).

Looking for your results ? Open one file and search for your name (as entered in your c2 logbook) or, better, search for your c2 LogbookID (a number, mine is 1528069, you can find it on your c2 profile page).

Additional details:

1. I currently consider only RowErg rower type (i.e. no slides, no dynamic). But I can easily change that to include the other types of rower (and take your best time if you have, let's say, entries for the 2K both on a RowErg and on Slides), if savvy people think that's better.
2. I only consider verified results and race results (i.e. no unverified results). Notice however that if your race result has been entered without your LogbookID, it will not count.
3. I've produced the results described above only for Rower events, but I can easily add SkiErg events (with same c2athlons, Core / SmallSplit / BigSplit / FullRange) and even mix them (e.g. Core / SmallSplit / BigSplit / FullRange on both Rower and SkiErg).
4. In each c2athlon file, sorting is by total score, with all the included events being mandatory. If you haven't logged a ranked workout for on included event you will have a TotalScore of zero. There's however another column (TotalScoreNME) in which the total is done without requiring a ranked workout in all events included in the c2athlon. Example: for TheCore you need a 2k a 5k and a 10k. If you only have a 2k and a 5k, your TotalScore will be 0, but your TotalScoreNME (non-mandatory events) wil be the sum of your 2k score and 5k score (plus a 0 for the 10k ).

Phew, long post. Any comment welcome.

[HornetMaX]

Season 2023-24 over, I've uploaded the results in the usual folder: https://drive.google.com/drive/folders/ ... sp=sharing