Page 22 of 24 FirstFirst ... 122021222324 LastLast
Results 211 to 220 of 238

Thread: Arena XII Tracking Thread -- Hi Everyone

  1. #211

    Default

    238,860 Hero

  2. #212

    Default

    Quote Originally Posted by Bill Bagot View Post
    The final prediction numbers are up guys. Thanks for all the data. Hope you all enjoyed the arena and stopping in here to see how it was unfolding. Hopefully next time Rachel will be in a place to be able to whip up one of her spreadsheets again. :-)

    I know what she does means a lot more than my pen and paper stuff.

    Hope everyone gets their desired ranks and "May the odds be ever in your favor" ;-)
    Bill. Thanks for starting Thread. Yes I did miss Rachel's spreadsheet (I don't think I'm alone there.) It was a very good Guide. Cheers to all that added total's too.

  3. Default

    Quote Originally Posted by Bill Bagot View Post
    The final prediction numbers are up guys. Thanks for all the data. Hope you all enjoyed the arena and stopping in here to see how it was unfolding. Hopefully next time Rachel will be in a place to be able to whip up one of her spreadsheets again. :-)

    I know what she does means a lot more than my pen and paper stuff.

    Hope everyone gets their desired ranks and "May the odds be ever in your favor" ;-)
    Thank you Bill.

  4. #214

    Default

    Quote Originally Posted by Dokkōdō View Post
    What you are describing is nothing new. It is called greed. It is not the model fault. It stems from human greed. That is all.

    You can see this in real life as well, there is a concept that is called, the "tail wags the dog" (or sth like that, it has been many decades since my finance grad school days). It is the relationship between the stock option and the stock itself. Predict indicators would drive up the underlying assets due to people expectation, and in turn, drive up the derivative markets as well.

    There are many other examples, such as the housing market bubble in the US, the Tulip market crashed February 1637 (Netherland), etc.

    If you are going to blame the spreadsheet for arena points inflation, then you might likely blame every single problem in the world as someone else fault. The problem does not lay in the world but within each person. If you are not inherently greedy or have some capacity to limit yourself then no problem.

    I saw this arena was just too insane and I bow out early, b/c I don't want to chase and follow everyone else. The spreadsheet created by some good folks in this forum is only a tool, which you can choose to use it or not. When things get too out of hand, people says F.IT and walk away, and the bubble will burst; the cycle will start anew. It is part of nature, things rise and fall through peaks and troughs. The duration of the cycles might varies, but the trend is basically the same for all things in nature.

    TLDR: Do not blame the spreadsheet model; do not blame the people who help make the spreadsheet; blame yourself if you are not capable of self-restraint.
    Actually, the derivative models lead the underlying asset class prices. That's why futures are called futures because they are pricing in future price and why people look towards unusual activity in the options market to determine if there is something that they have missed in the underlying asset and then the computers come in if there is an arbitrage opportunity. But that's beside the point. When using predictive models, all data points should be considered and then determine which ones are the outliers and then adjust the models to exclude those outliers. In your example of finance, recall the flash crashes and how some stock prices were well out of range of of even a two or three standard deviation move. These moves are such outliers that they are ignored when using predictive stock pricing models. As all data points aren't available in arena and the reason for the outliers aren't known they shouldn't be included in the arena prediction model. But as CA's computers don't know who is just doing so much extra just to get it done. their rankings can't distinguish and doesn't explain the moves. However, the model created in the spreadsheet doesn't factor this in because it can't as they don't know the reasons either. The predictions the spreadsheet is making is based on incorrect or flawed data which is why it overinflates estimates. If you know anything about numbers even a minor error at the beginning will cause and major distortion at the end. If you use a compass and walk one degree in the wrong direction, you can end up miles from the final location. One degrees is a small fraction 1/360 vs the initial assumption errors in the spreadsheet. It doesn't account for the fact that 50 tokens are given at the start, but in reality there aren't 28 hours in a day which is what is assumes. Plus it doesn't account for the fact that the first rankings are out more than 24 hours after the start of arena. In this arena, it started at 3 AM eastern time, but the first rankings were publish 25 hours later at 4 AM eastern. So the spreadsheet is making predictions on a 29 hour day instead of a 24 hour day.
    Last edited by John944615; 12-28-2018 at 12:06 PM.

  5. #215
    Join Date
    Aug 2009
    Location
    California, USA
    Posts
    5,141
    Blog Entries
    4

    Default

    Quote Originally Posted by John944615 View Post
    it clearly points out why the daily prediction model and the daily changing predicted final AV score needed model is bad for arena and why it inflates scores.
    Only 5% of players would care about the final AV score.

    For people in the 4-5% range, the predicted AV line encourages people to stop trying. The spreadsheet information about the AV line probably keeps the Legend line from getting too far out of control, and that in turn helps all of the people in even lower ranks.

    So, the spreadsheet helps those of us in the bottom 96% of Arena. People in the top 3% should be able to take care of themselves.
    Just another Castle Age player
    army code: D422AE

  6. #216

    Default

    Quote Originally Posted by Falco View Post
    Only 5% of players would care about the final AV score.

    For people in the 4-5% range, the predicted AV line encourages people to stop trying. The spreadsheet information about the AV line probably keeps the Legend line from getting too far out of control, and that in turn helps all of the people in even lower ranks.

    So, the spreadsheet helps those of us in the bottom 96% of Arena. People in the top 3% should be able to take care of themselves.
    I'm sure everyone know how to add and has access to a calculator. Anyone in the bottom 95-96% can do their own estimates, and also because the numbers are lower for reaching a certain rank, the errors that I talked about won't affect it in the spreadsheet. Are people going to put up 500,000 points on the first day to finish in gladiator, hero, or legend? No. That is why it doesn't affect the lower rankings to the extent it effects AV and EV ranks.

  7. #217

    Default

    Quote Originally Posted by John944615 View Post
    Actually, the derivative models lead the underlying asset class prices. That's why futures are called futures because they are pricing in future price and why people look towards unusual activity in the options market to determine if there is something that they have missed in the underlying asset and then the computers come in if there is an arbitrage opportunity. But that's beside the point. When using predictive models, all data points should be considered and then determine which ones are the outliers and then adjust the models to exclude those outliers. In your example of finance, recall the flash crashes and how some stock prices were well out of range of of even a two or three standard deviation move. These moves are such outliers that they are ignored when using predictive stock pricing models. As all data points aren't available in arena and the reason for the outliers aren't known they shouldn't be included in the arena prediction model. But as CA's computers don't know who is just doing so much extra just to get it done. their rankings can't distinguish and doesn't explain the moves. However, the model created in the spreadsheet doesn't factor this in because it can't as they don't know the reasons either. The predictions the spreadsheet is making is based on incorrect or flawed data which is why it overinflates estimates. If you know anything about numbers even a minor error at the beginning will cause and major distortion at the end. If you use a compass and walk one degree in the wrong direction, you can end up miles from the final location. One degrees is a small fraction 1/360 vs the initial assumption errors in the spreadsheet. It doesn't account for the fact that 50 tokens are given at the start, but in reality there aren't 28 hours in a day which is what is assumes. Plus it doesn't account for the fact that the first rankings are out more than 24 hours after the start of arena. In this arena, it started at 3 AM eastern time, but the first rankings were publish 25 hours later at 4 AM eastern. So the spreadsheet is making predictions on a 29 hour day instead of a 24 hour day.
    Great, you say a lot of stuffs on more than one occassions to show us how smart you are.

    Now the real question is:

    Are you going to make a better spreadsheet model and share with everyone here in the forum for the next arena? Are you going to actually DO ANYTHING or just talk?

    If not, just say thank you to the people who put their efforts into helping others with some form of tool, however simplistic or flawed. If you don't like the tool they provided, don't use it, don't open it or look at it. End of story.


    Nemo vir est qui mundum non reddat meliorem.


  8. #218

    Default

    Quote Originally Posted by Dokkōdō View Post
    Great, you say a lot of stuffs on more than one occassions to show us how smart you are.

    Now the real question is:

    Are you going to make a better spreadsheet model and share with everyone here in the forum for the next arena? Are you going to actually DO ANYTHING or just talk?

    If not, just say thank you to the people who put their efforts into helping others with some form of tool, however simplistic or flawed. If you don't like the tool they provided, don't use it, don't open it or look at it. End of story.

    The spreadsheet we use is fine. The problem is the predictive part of it. Eliminate that part and post actual crossover numbers and it's simple math and people can see if the rankings are moving a certain amount of points every day and then if there is one day it moves above the range and then falls back into the range, it will continue on the same trajectory. Using the 3 day moving average doesn't account for a deviant day but factors it into future predictions. I did make one in the past, but without the predictive part of it just used historical numbers, but since it didn't have the predictive part of it, people preferred not to use it. But it did show historical hourly rates and did factor into the additional tokens given at the beginning and the more than 24 hour first ranking and showed what the current points were needed at that specific time for a certain ranking based on the historical hourly averages so people didn't have to wait for the updated rankings to find out where they were.
    Last edited by John944615; 12-28-2018 at 12:30 PM.

  9. #219

    Default

    Very difficult to play Arena XII when I only got access to the feature 3 days ago. The Arena XII icon appeared one day, however, I couldn't access the feature. Then it vanished, and reappeared 3 days ago.

  10. #220
    Join Date
    Jun 2010
    Location
    Sacramento, CA
    Posts
    173

    Default

    I'll second that. All your hard work is greatly, greatly appreciated.

    We all benefit, and we should all be thankful.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •