S&P 500 weekly competition for 2012

[robster970]

Re: W/E 11th May - Results

However, one suggestion I'd like to make for the future which could be added in to competition is to do a distance based scoring system. So the idea would be that the winner of the quarter would have the lowest cumulative amount of points at the end of the quarter.

I too have opinions about the scoring mechanisms but I did model your suggestion just out of curiosity on the weekly score sheet, some time ago.

https://docs.google.com/spreadsheet/ccc?key=0AnFF2Rblu36wdGRNQjNXTzVjdldJb2dXdjRVTGs5VVE#gid=2

You will see a tab called group stats and your suggestion of the cumulative (also taking into consideration whether people are + or - too) demonstrates that tar, who participates infrequently would win. My problem with this is that regular participants are penalised because their increased sample sizes demonstrate their long run forecasting capability. I am not sla@gging off tar here btw but he does not participate that frequently and therefore I believe his win is a statistical aberration rather than relating to a long running statement of performance.

So before trying to improve the scoring system, perhaps we should question what the purpose of the weekly comp is. Is it:

a) Demonstrate accuracy on a predictive basis for the week, alone (in which case the cumulative suggestion would be a good scoring system)
b) Represent a longer run forecasting capability over say a quarter or year (in which case the current or Pat's proposal would better suit).
c) Something else.

Food for thought lads.

 

[Atilla]

Re: W/E 11th May - Results

I too have opinions about the scoring mechanisms but I did model your suggestion just out of curiosity on the weekly score sheet, some time ago.

https://docs.google.com/spreadsheet/ccc?key=0AnFF2Rblu36wdGRNQjNXTzVjdldJb2dXdjRVTGs5VVE#gid=2

You will see a tab called group stats and your suggestion of the cumulative (also taking into consideration whether people are + or - too) demonstrates that tar, who participates infrequently would win. My problem with this is that regular participants are penalised because their increased sample sizes demonstrate their long run forecasting capability. I am not sla@gging off tar here btw but he does not participate that frequently and therefore I believe his win is a statistical aberration rather than relating to a long running statement of performance.

So before trying to improve the scoring system, perhaps we should question what the purpose of the weekly comp is. Is it:

a) Demonstrate accuracy on a predictive basis for the week, alone (in which case the cumulative suggestion would be a good scoring system)
b) Represent a longer run forecasting capability over say a quarter or year (in which case the current or Pat's proposal would better suit).
c) Something else.

Food for thought lads.

As long as any system is applied equally to all participants I don't mind which system we use. I also concur that major game rules should only be changed at start of new quarter.

However, if an entrant plays in one game and not others then they should get nil points for non entries. You've got to be in it to win it n all that. Players conform to game rules not other way round.

Game objectives as I understood it was to call weekly and get direction and level as close as possible to Friday close. Clear enough for me.

I'm not fussed about points or how close one is. I'm usually happy to be in the direction of trend and manage the trade to get +ve pips scaling in or out.

Robster is also well advanced in Excel to be able to calculate pips made or lost too but without stops and limits what is the point of changing scheme to points basis. Not going to make much difference imho. However, once again I don't mind trying out new scheme next quarter if players choose to do so.

 

[hwsteele]

If I were in charge...I would run things differently!

Isn't that always the way?
Like I said before I will keep playing however the scoring is done.

The truth of the matter is that I would keep multiple data sets that would score the performance of everyone, each data set, differently.
From this I would then compile a total score based on the score of each section of performance.
The placement of the people (1st, 2nd, 3rd, and so on) would be based on the total score of all data categories combine.

The reason for doing this is best demonstrate by the fact that there are multiple ideas given on how to keep score. I believe the reason for this is that people look at the market in different ways. For instance, the idea of direction being important is a general truth.
BUT
Not from Friday to Friday, but from swing to swing. I don't know too many people who guess a direction for the whole week and then enter a trade on Monday and only exit on Friday.
To me the person who gets the closest to the weeks close number should win that particular contest regardless of direction for the week. Getting the direction right could be an extra point, but you should be given points even if the direction is not correct.

To me the contest is more along the lines of placing a pin on a number at a point in time. To do this you have to take into account the possible divergences from the open to the close(of the week)of the market. If the market moves 50 points against you only to come back 48 points and you still lose (get no points) because you missed it by 2 points in the wrong direction...that just seems wrong.
You could be within that 2 points and the next person was 10 points away, but in the right direction. You were closer but the other guy still wins.

There are so many ways to keep score that it might better represent the feat being accomplished by keeping score more than one way and then combining.

But this is the nerdy statistic loving guy talking here. keeping things simple is also a great way to go!

All IMHO of course.:innocent:



P.S. Robbie boy rocks for doing all this in the first place so I am happy to have him make the final choices. Maybe a little more input from Godfather Pat would be good too, or a little more weight from said input.

 

[ibetyou]

1360 for me please.

 

[robster970]

I am pretty relaxed on making a change. Atilla is right, I am pretty handy on excel so doubt that you are likely to come up anything that I can't do and make simple to administer. My only caveat is that we change at the beginning of Q3 if people want to change.

Based upon what I think people are saying, what about something like this:

a. People would get 1pt for participation
b. Podium continues
c. I keep tabs on the cumulative differences over a quarter per entrant

The leaderboard is constructed by dividing the cumulative difference by the points accumulated. The person with the smallest number wins at the end of the quarter.

This mechanism therefore:

i. Rewards participation - the more you play the greater your divisor becomes - this rewards participation
ii. Podium rewards accuracy for the week - again the divisor increases with weekly accuracy prowess
iii. Dividing cumulative difference by number of points rewards directionality as well as long run performance.

All criticism welcome, just throwing this out there.

 

[Pat494]

I am pretty relaxed on making a change. Atilla is right, I am pretty handy on excel so doubt that you are likely to come up anything that I can't do and make simple to administer. My only caveat is that we change at the beginning of Q3 if people want to change.

Based upon what I think people are saying, what about something like this:

a. People would get 1pt for participation
b. Podium continues
c. I keep tabs on the cumulative differences over a quarter per entrant

The leaderboard is constructed by dividing the cumulative difference by the points accumulated. The person with the smallest number wins at the end of the quarter.

This mechanism therefore:

i. Rewards participation - the more you play the greater your divisor becomes - this rewards participation
ii. Podium rewards accuracy for the week - again the divisor increases with weekly accuracy prowess
iii. Dividing cumulative difference by number of points rewards directionality as well as long run performance.

All criticism welcome, just throwing this out there.

iii seems to reward directionality equally or am I missing some point ? This would seem to benefit those that always made tight ( near the last Friday's close ) calls too.

 

[Pat494]

I would love to mention to Bernie Ecclestone just how bad the F1 scoring system is. It could be better designed to bring car racing back to life. But that's another story I guess.

 

[robster970]

iii seems to reward directionality equally or am I missing some point ? This would seem to benefit those that always made tight ( near the last Friday's close ) calls too.

I've just done a bit of fiddling with the actual sheet. Having a cumulative of the diff doesn't work too well but what would work well would be to have 1pt for direction as we currently do and then the cumulative absolute diffs added.

This means the divisor increases with directional prowess (a good thing) and the numerator (the cumulative diff) does reward those who predict closer to the close.

 

[robster970]

I've expanded the sheet out a little with some other things I calculate. If people want to download, fiddle with it a come up with some suggestions then they are more than welcome to.

 

[Pat494]

We have a month and a half to discuss this, so let's not rush it imho

 

[Pat494]

NEW YORK (Frankfurt: A0DKRK - news) , May 11 (Reuters) - Shares of U.S. banks slumped on Friday after JPMorgan said it lost billions of dollars on bad trades, but the overall market ended only modestly lower, thanks to gains in technology shares.

JPMorgan Chase (Hanover: 850628 - news) & Co, the largest U.S. bank by assets, dropped 9.3 percent on record high volume after it disclosed losses on derivatives trades. The news sparked fears that the problems could reverberate through the banking sector. The KBW bank index fell 1.2 percent.

"JPMorgan will become a political issue. This will increase regulations on banks, and the overhang on large banks will last for awhile," said Tim Ghriskey, who oversees about $2 billion as chief investment officer of Solaris Group in Bedford Hills, New York.

JPMorgan's shares fell to $36.96, and trading volume was around 216.7 million shares on Friday.

Wall Street ended lower for the second week in a row, as concerns about Europe (Chicago Options: ^REURUSD - news) 's fiscal health resurfaced as political turmoil in Greece again sparked worry that it could exit the euro and Spain's ailing banks spurred fears the country could need a bailout, while some U.S. economic data raised questions about growth.

But a survey released on Friday showed U.S. consumer sentiment rose to a more than four-year high in early May as Americans remained upbeat about the job market. The survey was a welcome sign amid worries that the economic recovery may be slowing down.

The Nasdaq finished higher, outperforming the broader market. Shares of chip maker Nvidia Corp rose 6.4 percent to $13.21 after quarterly results beat expectations. The stock boosted the Nasdaq and was the S&P 500 (SNP: ^GSPC - news) 's top percentage gainer.

Other chip stocks also rose, with the Philadelphia Semiconductor index up 0.73 percent. Also in the tech sector, shares of Netflix (NasdaqGS: NFLX - news) jumped 6.8 percent at $77.38.

The Dow Jones industrial average was down 34.44 points, or 0.27 percent, at 12,820.60. The Standard & Poor's 500 Index was down 4.60 points, or 0.34 percent, at 1,353.39. The Nasdaq Composite Index was up 0.18 points, or 0.01 percent, at 2,933.82.

Chesapeake Energy Corp shares fell 13.9 percent to $14.80 after the company said it may delay assets sales in order to preserve cash flow needed to comply with requirements of its corporate credit facility.

For the week, the Dow fell 1.7 percent, the S&P fell 1.1 percent, and the Nasdaq was off 0.8 percent.

Marc Pado, a U.S. market strategist at DowBull.com in San Francisco, said traders had helped support the market by closing short positions - bets that stocks will fall - after gains at the start of May.

"The trader types see that we came down to that 1,340 area on the S&P 500, started to bounce, started to see some buying, some bottom fishing, then you got that consumer sentiment number, and that was compelling enough," he said.

The disclosure by JPMorgan came as shocking news by a bank viewed as a strong risk manager.

JPMorgan estimates the business unit involved in the trading loss will lose $800 million in the current quarter, excluding private equity results and litigation expenses. The bank had previously expected the unit to post a profit of about $200 million.

Jamie Dimon, JPMorgan's chief executive, cautioned that losses could grow by another $1 billion, another hurdle for a sector already besieged by the sovereign debt crisis in Europe and fears of slowing growth globally.

The news weighed on bank shares as investors feared both a greater risk of more regulation and the potential for more such losses at other banks. The stocks, however, came off their lows of the morning.

Citigroup Inc (NYSE: C - news) lost 4.2 percent to $29.35 and the Financial Select Sector SPDR was off 1.1 percent at $14.82 and the S&P financial sector fell 1.2 percent.

Financial stocks have been among the most volatile in recent months as investors question what the growth outlook for the United States and the European debt crisis will mean for the group's profits. JPMorgan has fallen 14 percent this month.

The CBOE VIX Volatility Index rose nearly 16 percent for the month in a sign of growing caution.

Thomson Reuters (Toronto: TRI.TO - news) /University of Michigan's preliminary consumer confidence index for May improved to a reading of 77.8 from 76.4 in April, topping forecasts of 76.2.

Of the 453 companies in the S&P 500 that have reported earnings to date for the 2012 first quarter, 66.2 percent have reported earnings above analysts' expectations, according to Thomson Reuters data. That compares with more than 80 percent at the start of earnings season and is below the average for the past 4 quarters of 68 percent.

Shares of Arena Pharmaceuticals Inc (NasdaqGM: ARNA - news) jumped 74 percent to $6.36 after a panel of experts recommended approval of the company's obesity pill, a big step toward making it the first new diet drug on the U.S. market in more than a decade. The stock was the most actively traded on the Nasdaq composite (Nasdaq: ^IXIC - news) .

Volume was about 6.47 billion shares on the New York Stock Exchange, the Nasdaq and the NYSE Amex, above the daily average of around 6.8 billion. (Editing by Leslie Adler)


I hope you have the time to digest this as it could be relevent.

 

[Atilla]

1332 for me please Rob...

Trend is down and likely to attempt the upside 1370s but market conditions poor and considering we are in May - guestimating up-down week.

 

[wackypete2]

1339

 

[Gaffs1964]

1373 for me please Rob

 

[samspade79]

1379

 

[bodavula]

1342 for me please!

 

[DonStar]

Re: W/E 11th May - Results

I like the above idea but I still think direction is important. After all it's no good from a trading point of view being near the end of the week's figures if one has the direction wrong.

Sometimes you can be within 1 point and have the direction wrong.

So I disagree.

1340 this week Rob...

 

[robster970]

Re: W/E 11th May - Results

Sometimes you can be within 1 point and have the direction wrong.

So I disagree.

1340 this week Rob...

New thread started to discuss and agree changes off the main comp thread

http://www.trade2win.com/boards/foyer/150230-scoring-s-p500-weekly-comp.html#

 

[robster970]

1376 for me this week - agree with isatrader, last week was institutional buyers putting in a long term bottom whilst retail was selling into them.

 

[VielGeld]

1380 please.