PPA seeding run already includes last race

[MCDreyer]

Worcester race already included in new seeding run....!!!

 

What's up with PPA this year?

 

After the "New Argus" seeding run everything seems messed up. Not that I complain about the last race being included already but things are just different and very strange to me. They normally only re-seed on the first Wednesday of the month.

[PeterM]

Hi Vaste6

 

I also found the new ppa seeding process confusing.

After contacting the ppa numerous times and ways ? and receiving no/zilch/nata response from them, I found the guy who does the actual calculations and the changes for this new season. He was very helpful and we had a long chat.

If you want to, I could try to explain it ? but please note that, what ever I will say is only 2^nd hand info, plus sometime in the future there will be an official ppa anoucment regarding it. But maybe it might help in the meantime.

Let me check my notes re the seeding and write them down - will poste them later on.

PeterM2009-09-25 02:08:19

[MCDreyer]

Thanks PeterM.

I don't mind riding where I am but would be great to know a bit more.

[PeterM]

OK, here it goes:

 

Apparent changes to the PPA seeding process for the new season:<?: prefix = o ns = "urn:schemas-microsoft-com:office:office" />

 

Please note: what I am now recalling is 2^nd hand information to you; so there might be a few mistakes that crept in, etc.

 

The basic calculation has changed. (old calculation = adjusted winner's time, divided by - your time ? converted into a %). Now it is: your time, divided by the winning time. This is mathematically a cleaner/neater way of doing it. Example: if your personal race time is 240min and the winning time is 200min = than your (initial) seeding is 240/200 (+convert to %) = 20 i.e. your time is 20% more than the winning time. Read on:

 

Than the winning time is adjusted. Similar as before to take the race circumstances into account. Follow on from the above example ? winning time is adjusted from 200min to 180min. 240/180 (+convert to %) = 33,3 (your adjusted race seeding)..

 

Now the ?spread? of the participants is taken into account plus other circumstances; i.e. front guys are racing seriously, cyclists at the back are just doing a social ride with their mates; or the stormy Argus race, etc. (There is a level of subjectivity involved) This is a type of linear regression. That adjustment might go either way - can be positive or negative!! Example - and your race specific seeding is again adjusted let?s say by 15% = 33,3 & 15% = 38,3 (your adjusted race seeding). 

 

4.   And now additional statistical info is taken into account, size of group (population), etc.

 

The weighting process seemed to be the same.

 

Practical effects:

It will be more accurate overall

But it will need continuous & more adjustments and fine tuning!!

How transparent will the adjustments be?

Personally, I think that the process was too strict and punitive for the <?: prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" />Worcester race. Looking at my own Worcester race seeding - it was terrible despite that I have been training during the whole winter. But then I applied the new calculations process to the league riders - I found out that many of them would also have received a terrible race seeding and many might not have ?qualified? for leagues. This means that the Worcester seeding might have been too strict!

There will be fewer races that give one an easy improved seeding; i.e. West Coast Express ? might not be the easy seeding-improving race anymore.

Small races will be sidelined ? too small population, etc

Bigger attended races might be better for once seeding.

Tough hill races will be fairer judged and be good for once seeding.

 

I hope that helped. Please note that there might be a few mistakes in it.

 

Peter M?ller

[udxcob]

Isnt the reseeding supposed to be done on the first wednesday of the month?????.....................................??????

[FanieFiets]

well, it is out if you want to have a look. 

[waynex]

In the new system a seeding for the Durbanville High Race that was 16.5 has now turned into a 19.5. ???? It just makes things a bit more diffucult.

[FanieFiets]

 

and from 9 to 11.5 for me. Interesting. It is a double adjustment and not transparent at all. 

FanieFiets2009-11-05 02:07:31

[Woofie]

What do you mean?

Was it 16.5 for you last year and now its 19.5?

 

 

[FanieFiets]

using the old formulae of 100-(adjusted winner/your time x 100) gives me a lower seeding that the one ppa gives. It must be the new formula peterm describes above, althou he lost me! 

[Tubehunter]

From the seeding site:

 

For events that took place before the previous funride season, ?penalties? will be added as these results are less current than events that took place in the last season. These ?penalties? are generally 5% per year.

 

Obviously the index from the previous years event will change as the same event transpires. That said a fair reflection of your seeding will require 3 current long rides from this years series. (Hopefully you had decent rides in all of them)

[Woofie]

 

using the old formulae of 100-(adjusted winner/your time x 100) gives me a lower seeding that the one ppa gives. It must be the new formula peterm describes above' date=' althou he lost me! 

[/quote']

 

I cant get the formula that he speaks of above to work either.

Or maybe I just dont know how to use the % sign on the calculator properly.

But I spent a whole evening going through my results and adjusted winners times and it baffles me how they work it out now.

 

Fanie, you are more connected with the ppa, cant you organise the new formula?

 

 

[wheelz]

WTF. . .how does my seeding drop from A to C????

[FanieFiets]

 

From the seeding site:

 

For events that took place before the previous funride season' date=' ?penalties? will be added as these results are less current than events that took place in the last season. These ?penalties? are generally 5% per year.

 

Obviously the index from the previous years event will change as the same event transpires. That said a fair reflection of your seeding will require 3 current long rides from this years series. (Hopefully you had decent rides in all of them)

[/quote']

 

talking about this race.  With last years formulae = 9, with the explanation on the previous page - 9.9 and on PPA's website 11.32.

 

That is taking into acount adjusting the winners time by 4 minutes.

 

I dont really mind as I do not want a lower seeding, but I do like to know how things work.

 

[Marius]

 

I love the new way PPA works the Seeding out, I have 3 x ZERO indexes that counts towards my seeding:-D That said, I still ended up with a 0.479 index, not sure were the figures after the zero came from..

 

Ok, they changed it since the last time i checked, I have a 2.5 now, damn

Marius2009-11-05 02:05:43

[waynex]

The new seeding calculation for the same race ae generally 2-3 point lower so I think most guys will drop down a group? The way i work out the new formula is   Adjusted winners time/your time X 100 - 100 = seeeding