Chart Game

Posted by golax on February 16, 2010 (11:21PM)

For those of you that would like to play a stock market game, you should check out the following link. It's fun!

ChartGame: http://chartgame.com/

Posted by incubus on February 16, 2010 (11:40PM)

That's just too cool, now if there was the ability to cost average in and out, I'd actually win once in a while.

Posted by don2277 on February 17, 2010 (12:23AM)

thanks G...great practice for the real world

Posted by golax on February 17, 2010 (01:31PM)

I played it for a while and was winning but I would never trade that way in real life. I'm not a chart person myself. But you can sort of test out the old MACD and a few other indicators in a crude way.
Img_4028_thumb Brokerage Account

Posted by 47 on February 17, 2010 (08:31PM)

Yea had quite a bit of fun with this today.  MACD seems to work very well, but I question how the effectiveness of MACD has held up over time as more people became aware of it/utilized it.  Any thoughts? 

Posted by affinity1 on February 18, 2010 (12:20AM)

Oh man thanks for finding this!!! So fun...

I did three charts and had an annualized return of +31% compared to the annualized buy and hold of 6%.

Too bad it doesn't factor in transaction costs.
Owl001_thumb Brokerage Account

Posted by Owl on February 18, 2010 (11:24AM)

That is an awesome game. Wish I'd found it before. Based on limited experience, it seems to me that the game shows that momentum trading works.
Owl001_thumb Brokerage Account

Posted by Owl on February 19, 2010 (05:20AM)

Only problem with the game is, it only includes stock market history going back to about 1990. I've written to the author and suggested including a longer history. Some trading strategies might be more successful in some time periods than in others.

Inc, if you're losing, just try doing the opposite. Whatever you're taking for a buy signal, take it as a sell signal instead.

The only bad news would be if, no matter what strategy you used, you wound up with the *same* results (on average) as buy-and-hold. That would show that the market is unbeatable.
Img_4028_thumb Brokerage Account

Posted by 47 on February 19, 2010 (07:55AM)

There are two things I would love:
1) additional indicators
2) the option to only play in a recent time period

As Owl mentions, certain strategies are likely to only be effective during certain time periods.  This is the perfect backtesting platform for me personally, because I can find something that theoretically or statistically works, and then test my willpower and execution. 

Posted by golax on February 19, 2010 (12:49PM)

47 said: There are two things I would love:

1) additional indicators
2) the option to only play in a recent time period

As Owl mentions, certain strategies are likely to only be effective during certain time periods.  This is the perfect backtesting platform for me personally, because I can find something that theoretically or statistically works, and then test my willpower and execution. 
 Yes...it all comes down to willpower and execution...the bane of my trading existence.
Owl001_thumb Brokerage Account

Posted by Owl on February 19, 2010 (04:13PM)

I wrote to the author, and he wrote back. He said that he didn't know where to get the data from before 1990. He said that if I could find it for him, he'd consider using it in the game. Anyone have any suggestions?
Owl001_thumb Brokerage Account

Posted by Owl on February 20, 2010 (06:21PM)

I'm surprised there is not more discussion of this. I've written a third of the messages in the thread. This game is such a great opportunity for testing charting techniques! Anyone who is an active trader should play the game. Multiple times. Further thoughts:

  • - Why is this game better than actual trading? (a) You can play it with no risk, (b) you can go through a year of trading in 20 minutes, (c) as a result of that, you can test your strategy out many times, to verify that its success is not due to chance, (d) you can test strategies on random time periods. This protects you against being biased towards strategies that would only have worked, e.g., during the past year (which was an unusual year).
  • - One problem is that you cannot switch from a long to a short position in one day. If you want to do that, you have to sell on one day, and then take another day to go short. Similarly with switching from short to long. This is a major handicap to a trend-following strategy.
  • - Another problem: you can only decide to buy/sell at the next day's open, without knowing what it is. In real life, you get to see the open before deciding whether to buy/sell.
  • - I wrote to the game author, pointing him to all the data on Yahoo. But I don't get the impression that he is interested in modifying the game.

Here is my experience. I post this, because I think we can learn from it:
  • - First, I tried the contrarian strategy: after the stock has just dropped, buy. After it has risen, sell. Result: buy-and-hold kicked my butt a couple of times in a row.
  • - Then, I tried the trend-following strategy: after a rise, buy; after a drop, sell/go short. Result: kicked buy-and-hold's butt.
  • - Then I played several more games, to verify that the result of the first few games wasn't just due to chance. Result: It was just due to chance. In a collection of 12 games, there was no consistent lesson. Sometimes the trend-following strategy won, sometimes it lost.
  • - I played several more games, trying different strategy variants. Result: I can't find anything that works consistently, such that I can say it isn't just chance.

My advice for active traders:

  • - Play the game.
  • - Don't just play it 3 times. Play it 10 or 20 times, all using the same strategy.
  • - Count how often you come out ahead of buy-and-hold. Unless you come out ahead at least 15 times out of 20 (or behind at least 15 times out of 20), your results are probably pure chance.
  • - If you do this, post your results. I am very interested in whether anyone can reliably beat buy-and-hold. (If you can't do it in the game, don't bet that you can do it in reality, even though the game is flawed as mentioned above.)

Posted by El Dorado on February 20, 2010 (11:52PM)

I'm surprised there is not more discussion of this. I've written a third of the messages in the thread. This game is such a great opportunity for testing charting techniques! Anyone who is an active trader should play the game. Multiple times. Further thoughts:

Owl, have you ever played Chess? Same board, but a bit more complicated than Checkers. 
Img_4028_thumb Brokerage Account

Posted by 47 on February 21, 2010 (03:34PM)

One I have tried quite a bit is using MACD and only using long positions (I do not feel comfortable shorting stocks at this point, even with four years of off and on trading experience).  Simple strategy: buy when the default settings indicate you should buy, and sell when default settings tell you to sell (i.e. buy line crosses below the sell line).  
I've had moderate success, though I am now going to try about 15 games with modified MACD settings (8-17-9).  I'll post my results when done. 
Img_4028_thumb Brokerage Account

Posted by 47 on February 21, 2010 (04:36PM)

Results were not good.  Anyone have interesting suggestions of technical combinations for the medium/long term trader?  Probably something for a different thread... I just get destroyed by false positives in the game.  

You must Log In to post to this forum.

Not a member? Register Now to …

  • See what other traders are doing
  • Make your own trades public
  • Share your thoughts on a trade
  • Join or start a group
  • Connect with like-minded traders