Use access key #2 to skip to page content.

My current accuracy rant (to link in pitches)

Recs

14

March 01, 2007 – Comments (19)

Sandworms!

Which is to say, you hate "accuracy", right? I know I do. Man, it's dumb. What's with the breakpoints? What's with the gains harvesting?

Ugh.

"Score leader" at least makes sense. Summing score linearly is dumb, but I guess it measures max-out-your-slots players against their 200-pick peers.

It's not that measuring how often someone is right is so bad, it's that the method is so gameable.

Measuring Goldman's picks this way could work. Measuring players who know 1/3 of their rating depends on the metric cannot.

19 Comments – Post Your Own

#1) On March 05, 2007 at 7:20 PM, sportscliche (99.60) wrote:

Welcome to the 99's, for whatever it's worth. You, turbotrager, and I assume others are clearly proving how gameable the scoring system is.

Report this comment
#2) On March 27, 2007 at 10:10 AM, AJL203PSU (< 20) wrote:

Please Stop with the accuracy rant. I've never "reloaded" a stock and i'm equal to you in rating. If you don't like how CAPS is set up, Close your picks out and go over to thestreet.com and play stockpickr.

Report this comment
#3) On April 20, 2007 at 9:13 PM, wcwhiner (33.70) wrote:

Funny, your CAPS rating (which I hurry to remind you I believe to be fatally flawed) currently is lower than that of my proof-of-concept player who makes random picks and ruthlessly reloads to cash in "accuracy".

I am curious, though: what precisely appeals about the "love it or leave it" line?

 

Report this comment
#4) On May 12, 2007 at 3:55 PM, ikkyu2 (99.18) wrote:

You know what, my current best pick is a +65.  It's just occurred to me that by sitting on this pick like a hawk and dividing it into 12 separate picks I could have boosted my accuracy by something like 10%.
You're right, this is dumb.  They should limit you to a boolean on every stock summed over all your picks: were you right more than you were wrong?  You get an accuracy point if you were; you don't if you weren't.

Report this comment
#5) On May 29, 2007 at 12:49 PM, Somnambulo (79.79) wrote:

What the game needs is an  list for accuracy on long term stocks only. Make it only count for stocks that a player has kept long enough to avoid capital gains tax for instance. That takes skill. Takes more skill than adding a pile of naked short candidates and waiting for them to burn to the ground. It also discourages people from reloading and playing the game like a swing trader. Underperfomers shouldn't be able to hold stocks indefinitely, or perhaps their score needs to separated out so that serious investors can find good long term picks based on skilled investors. I understand this game is not supposed to reflect the way the actual market works, but it needs a little tweaking.

Report this comment
#6) On June 07, 2007 at 5:29 PM, jonmchan (98.74) wrote:

somnambulo makes great points! I agree with him that this game does invite shorters and day traders and doesn't really put good long term fools in the lead. The top 100 are just people working the game system and although some are pretty skilled, I highly doubt that you can run your real portfolio the way they're shorting every other stock.

Report this comment
#7) On June 08, 2007 at 10:52 AM, ThePessimist (99.40) wrote:

A parallel issue is that for shorts you get more points for reloading. Say you have a stock at $5 and it goes all the way to zero. For simplicity, assume the S&P 500 stays stable. If you give the stock a single thumbs down, you get 100 points. But suppose you instead reload when the stock reaches $4.75 (giving you 5 points), and again at $4.51, and so on down to $1.50. You would pick the stock 24 times, on 23 of which you'd get 5 points and on the last of which you'd get 100, for a total of 215 points!

Of course you'd never get it that perfect in real life. But the point remains that on shorts, reloading improves not only your accuracy but your score as well. By contrast, reloading hurts the score for longs. That's too bad, since it'd be nice for CAPS to reflect real portfolios, which are generally long. (And which certainly don't have short positions in penny pink sheet stocks!) 

Report this comment
#8) On June 18, 2007 at 5:13 AM, leohaas (31.65) wrote:

Grow up, all of you! This is only a game.

If it really bothers you: It is easy to ignore the reloaders. Just compare the number of active picks with the number of ended picks, and draw you own conclusions!

Report this comment
#9) On June 20, 2007 at 8:40 PM, ez2foome (< 20) wrote:

I agree.  If you are posting the same stocks that are in your portfolio and you are scoring reasonably well, who cares about somebody with a higher score?

EZ    >^..^< 

Report this comment
#10) On June 21, 2007 at 3:05 AM, ajfabb (96.56) wrote:

This is a good reminder that CAPS scores should be taken in context and fully understood.  There is plenty of data on each player to understand what they are doing.

Report this comment
#11) On June 27, 2007 at 11:24 AM, jyurow (54.42) wrote:

In the context of comments from leohaas (#8) and ajfabb (#10), I guess my "buy and hold" game strategy pretty much mirrors my portfolio and fits with what ez2fooome (#9) writes. 

Report this comment
#12) On June 27, 2007 at 10:35 PM, majakblue (< 20) wrote:

I am using CAPS to learn about investing, and I am indeed learning a lot.   Meanwhile, I have a few stocks in real life that are a small percentage of my CAPS picks.  I will not do any substantial investing IRL until I am convinced that I have learned a lot more than  I know now.  Assuming that there are a lot more players like me, the CAPS ratings on stocks will be less important than knowledgeable comments per ajfabb's comment.

Report this comment
#13) On July 19, 2007 at 4:48 PM, Gumfactor (47.55) wrote:

I think it's actually worth noting that there is an accuracy/points trade-off to "reloading", such that the reloader is likely to boast a higher accuracy with fewer points. And since the points account for 2/3rds of the CAPS score, the reloader may lose out in the end.

Let me provide some numbers:

Consider "Investor", who buys stock A for $5.00, and watches this stock rise to $10.00. Voila: Investor A has a 100% profit, or 100 CAPS points.

Now consider "Reloader" B, who buys stock A for $5.00, sells at $6.00, buys at $6.00, sells at $7.00, buys at $7.00, sells at $8.00, buys at $8.00, sells at $9.00, buys at $9.00 and sells at $10.00. In the real world, Reloader would have made the same 100% profit as Investor (albeit with more transaction costs). But in CAPS, Reloader would have made 20% on the first sale, 16.7% on the second sale, 14.3% on the third sale, 12.5% on the fourth sale, and 11.1% on the fifth sale...for 79.6 total CAPS points. 

And so, although Reloader would be 5 for 5 on accuracy, he's have gained only 79.6% of the CAPS points that Investor would have amassed. And, again, since accuracy only counts for 1/3 of one's rating, the loss in CAPS points is very like to keep Reloader's rating similar to what he'd have otherwise.

Or at least that's the way I see the numbers.

For the record, I'm not a reloader and am not condoning such practices. But I don't know that they're creating the hienous artificiality within the CAPS world that blogs like this are insinuating.

Feel free to correct my numbers in any way one sees fit.

Matt

Report this comment
#14) On July 24, 2007 at 4:42 PM, Gtrinvestor (99.75) wrote:

I used to be so/ so on your accuracy rant, until I noticed how far my ranking dropped after a few accuracy % points went away.  If you compare my total points vs yours or even more importantly points growth rates, you would think I would be rated far ahead of you, or at least close.  Instead, you are rated far ahead due to your "accuracy", even though most of my "out of the money" bad picks are just slightly out of the money, and the in the money ones are way in the money.   

Now I am turning into an ugly point harvesting fiend, which I am sure was not the original intent of this tool.  Oh well, I still think a lot of the insights in CAPS is great, but the points system needs to get an overhaul... sigh.

By the way, nothing personal in regards to your points, I know it got killed when you made the big bet against all things oil related (and I did the opposite), thus the inversion of our score charts.  Otherwise, I agree with a lot of your stock viewpoints, so keep rating and posting, as it helps me to buttress any true money investing decisions I do make.

Report this comment
#15) On August 01, 2007 at 4:16 PM, lquadland10 (< 20) wrote:

As a newcomer to the game I think of it as a way to learn about stocks. Some are a hold and some are in and out. Great way to learn as you go. Thanks for all the great insights. Rock on.

Report this comment
#16) On September 26, 2007 at 1:41 AM, hall9999 (99.32) wrote:

  I don't understand why TMF allows this madness to continue.  Reloading cannot help.  It can only hurt. 

  In real life, it doesn't earn me one wooden nickel more (in fact, it loses me money because of trading costs).  It does not change my portfolio, it does not change my tax consequences, and it does not reflect any change at all in my view of the company.

   In CAPS - ok, it does help the individual player game the system.  Other than that, it hurts them because it takes them away from doing real work and learning to be a better investor as well as decreasing their value to other players.  In addition, it helps the crud rise to the top.  As a consequence, some players who might otherwise be all-stars are pushed lower.  This devalues their influence on stock star ratings and lowers their perceived value to other players.

Report this comment
#17) On October 02, 2007 at 5:31 PM, VasilGrozny (43.83) wrote:

I only have 12 active picks, yet I regularly see my rating swing wildly b/w 50 and 75, sometimes within the same day, even though both my total and average pick scores have been pretty much on a steady climb for the past 2 months or so.  All because I happen to have 5 or so stocks hovering around 0 -- sometimes they're green, sometimes red, yet they never add up to more than 5-10% of my total either way.  Ridiculous.

Report this comment
#18) On January 01, 2008 at 4:54 PM, bluejonnyd (22.71) wrote:

This system works best if it is not treated like a game.  Trying to get ahead in the rating system is the root cause of trying to 'game' the system.  IMO, the best way to fix it would be to place less emphasis on the main page on player preformances, and more on the contests.  This would remove some incentive for people to toy with the published ratings, as they would be doing it in the contest space.  More importantly though, the focus of players shouldn't be so much about trying to best one another, but instead be on contributing to the tool that is CAPS.

Report this comment
#19) On September 18, 2008 at 4:46 PM, djangophile (67.86) wrote:

A BETTER WAY: LOGs

I'd suggest that CAPS add a new rating number for each player, and leave the old one as is.  The LOG of each individual gain or loss should be totaled to create the score.  (As opposed to the percentage.)

For those familiar with Decibels, it would work just like that.  If something doubles, it goes up 6 dB.  if it gets cut in half, it goes down -6 dB.  If it gets cut in half again, it's down another 6 dB.

 That way, it wouldn't give any more points to reloading on the way down.  It would also remove the penalty for reloading on the way up.

And accuracy should only count for the current picks.

Report this comment

Featured Broker Partners


Advertisement