Run-Down Forums

Dan · Posted: Tue Oct 23, 2001 2:03 pm Post subject:

Here are a few more possible meets to fill in the gaps (there is a full month between Herculis and Weltklasse) and boost us up to 10 meets or however many we choose to go with:

Athens Grand Prix - Athens, Greece (mid june)
Zagreb Grand Prix - Croatia (early july)
Nikaia Grand Prix - Nice, France (early july)
DN Galan Grand Prix - Stockholm, Sweden (mid july)
Norwich Union British Grand Prix - London, England (late july)
IAAF Grand Prix Gugl-Meeting - Linz, Austria (mid august)
Rieti Grand Prix - Rieti, Italy (early september)

The calendar dates are from 2001, so no point posting exact dates for the coming summer... Some of them may be possible to find by going to the meet web sites, but not all of them have sites that I know of.

Dan

Conway · Posted: Tue Oct 23, 2001 6:48 pm Post subject:

Athens, Stockholm and Nice are all nice meets ... London is always cold and there are usually defections ... Rieti usually has good sprints ... They would make a nice addition to the schedule ...

Dan · Posted: Tue Oct 23, 2001 7:52 pm Post subject:

Athens, Rieti, Nice, and Stockholm were the ones that jumped out to me, too. Of course, that would make 11, killing our nice even 10, but I don't see that that actually makes any difference one way or the other.

Here's the ugly part of the Golden League schedule:

6-28 ExonMobil Bislett Games
7-05 Meeting Gaz de France
7-12 Golden Gala
7-19 Herculis Monaco
8-16 Weltklasse Zurich
8-30 Memorial Van Damme
9-06 Istaf Berlin

Note the amount of time between Monaco and Zurich, and according to last year's schedule, there isn't much going on in between other than the London meet, with Stockholm just catching the beginning of the gap. Of course, it's quite possible that some of those Grand Prix meets have been moved around to account for this, so I'll have to double check. At least we have a good list of meets to work with now.

That pretty much just leaves figuring out how to price these darn athletes for roster selections... Sad

I think everything will fall into place once we get that taken care of. I have the bulk of the program put together for running the league -- roster selection, scoring, and player leader board. The two major time items will be entering all the selectable athletes (with event and purchase cost) into the database initially, then entering the results for each meet. Everything else should be reasonably automated.

Dan

Dan · Posted: Tue Oct 23, 2001 10:36 pm Post subject:

Unless any of you can find dates on the following event sites, it appears no 2002 dates have been selected/posted yet:

http://www.tsiklitiria.org/ (just found this one)
http://www.dngalan.com/
http://www.nikaia.org/
http://www.rietimeeting.com/

Dan

Conway · Posted: Wed Oct 24, 2001 8:12 am Post subject:

Hmmm ... Pricing of athletes ... I'm going to just throw out something simple that we can compliate as we choose .. Smile

.. My thought is this, lets start by using the IAAF tables/rankings ... Assign a value of say 1000 pts to the #1 ranking ... Then lets scale down from there - say 100 pts per place ... Would give us 10 athletes per event ..

Could then a purchase maximum of say 60% of total events that we choose to have ... So, for example, if there were 10 events then the max purchase points would be 6000 .. Would ensure that no one could load up on #1 ranked athletes ... And would force some startegy in team selection ...

Just a thought off the top of my head ...

Micah Ward · Posted: Wed Oct 24, 2001 3:03 pm Post subject:

I like Conway's purchasing plan...it is simple. I am a great proponent of simplicity.

Dan · Posted: Wed Oct 24, 2001 3:59 pm Post subject:

I knew Micah would like the simple suggesetion! Wink

However, as simple as it may be (although I'm not sure I actually agree that it is any simpler...), is it any more accurate, or even as accurate? It creates more point separation between the athletes in the ranking (same idea behind my suggestion of removing 1000 from the IAAF ranking to yield values betwee 0 and 480), but I don't see there being any intrinsic meaning to 100 point gaps. Are Allen Johnson and Anier Garcia really separated by 10% in value???

As far as why I think it may actually be less simple, it requires us manually specifying the 1000-900-etc point values (admittedly, a pretty darn easy thing to do), whereas with the "subract 1000" method it's just a matter of going down the column of points in the text file version of the lists and deleting a character (1), or creating a regular expression to do it for the 50 names in each list with a single click. Smile

Dan

Dan · Posted: Wed Oct 24, 2001 4:07 pm Post subject:

I just noticed something interesting on the IAAF rankings lists -- they also indicate much lower scores on average for the sub-WR throws performances, especially when looking at the 100-deep lists.

Dan

Conway · Posted: Wed Oct 24, 2001 7:50 pm Post subject:

Like I said, I threw out the simple knowing it would need to be complicated some ... Smile

... I agree that a 10% separation may be too much ... So have no problem with the take away the 1 theory ...

So how to conduct the draft ???

Dan · Posted: Thu Oct 25, 2001 11:12 am Post subject:

Well, what I've got 99% put together and 90% polished (or something like that) is a list of the top 100 athletes in each of our selected events, with their IAAF adjusted price and a maximum allowable to spend however each player sees fit on their roster. Aside from the pricing still not seeming entirely right, this system allows everyone joining in the game to select their roster however they want, regardless of what other players are doing. No selection order required!

How's that for simplicity, Micah? Wink

When the first meet rolls around, the rosters will be locked down and no more changes can be made. Up till then, you can do all the tinkering you want. Spent 398 points on Szelezky and realized there are two 200 pointers that you think will take the circuit by storm this year? No problem!! Just log into your account and make the change.

I haven't looked at how the numbers would play out, but here's a thought I had for how we could make the pricing a bit more accurate. Subtracting 1000 from the IAAF points gives us numbers between 100 and 450 for the most part. Most of the athletes are separated by about 10 points (on either scale; bigger percentage gap on the adjusted lower scale). Considering a win and a good mark in any of our 10 or so meets is worth 180 points or so, that 10 point purchase gap is not very significant... Seems to me that we should be doing something along the lines of multiplying the points for the lower scale by the number of meets, as that's basically saying take the odds of each athlete scoring well and multiply it by how many chances they have. Make sense so far?

So, Maurice's 448 points on the low scale would become 4480, compared to 3360 for Bernard Williams and 2020 for Terrance Trammell. Definitely makes you think twice about the value of the top tier athletes, but you also have to keep in mind that loading up on bottom rung folks may not net you any points in return... Wink

How's that sound?

Of course, that's after saying Conway's 10% separation was too much... Smile

However, this method at least takes into account their separation on the original IAAF scale, not just the order, theoretically making for somewhat more accurate pricing gaps.

Dan

[ This Message was edited by: Dan on 2001-10-25 15:13 ]

Conway · Posted: Thu Oct 25, 2001 11:36 am Post subject:

Sounds good to me ... Makes it very competitive and strategic ... A bit more than just drawing some names and seeing how they perform .. Smile

Dan · Posted: Fri Oct 26, 2001 7:17 am Post subject:

Yep! And it provides plenty of opportunities to second guess and kick yourself down the road. Smile

Dan

Conway · Posted: Fri Oct 26, 2001 1:22 pm Post subject:

Well ... Must be time to recruit !!!!

Dan · Posted: Fri Oct 26, 2001 1:52 pm Post subject:

A timely post, I must say... I have updated the price lists with the latest IAAF converted scale, and as best I can tell, the program is fully functional. So, time for you all to play with the new toys! Smile

http://run-down.com/fantasy/

Micah, Conway, and Justin -- I set each of you up with a user account with the same username/password combo as you have here for the forum. Should make it easy to remember, but you can change your password any time you want. For everyone else, I have the program set up to not require admin validation of user-created accounts, so feel free to get started and try things out.

If anything doesn't seem to be working quite right or is less than self-explanatory, let me know and I'll look into it. Now is the time to work out any kinks... If initial testing goes well, then we should have no problem being ready to go for a trial run come indoor season.

Dan

p.s. I set up two rosters -- "Dan Kaplan" and "Test User" -- for you to see some different roster selection approaches. One is rather top heavy, the other consists largely of cheap athletes, making for a drastic difference in roster size. There are only 2 or 3 sample results in the database from 2 meets, so there will not be many scores showing up in the leader board display.

[ This Message was edited by: Dan on 2001-10-26 17:56 ]

Micah Ward · Posted: Sun Oct 28, 2001 11:58 am Post subject:

OK guys, I have selected my roster. Some of the picks are good ones and some were made more with the heart than the head. Now I have a question. I selected Szabo (imagine that). If she runs the 1500 and the 3000 in the same meet do I get points for both? I selected her from the 1500 list.

LET THE GAMES BEGIN!!!!!

By the way guys, How's the real world training going? I bought new shoes last week and the foot is feeling fairly decent.
Micah

Discussion Forum

Share the Love!

What's New