Message Topic

New Analysis Monster page and Trade Monster updates
avatarken_admin (1534 posts)  9/16 5:12 AM

We’ve added an Analysis Monster page which simplifies the analysis of Projected Standings and Team Analysis.  It contains:

  • An updated Projected Standings better taking into account IR players and game limits.  We suggest using this rather than Projected Standings or Standings Monster though both of those are still available.
  • An updated Team Analysis using the stats from Analysis Monster.  We suggest using this rather than the old Team Analysis page though that’s still available.

We’ve also simplified the Trade Monster page which now uses the Analysis Monster results.  It allows:

  • Viewing Before/After Projected Standings & Team Analysis for Trades.
  • Viewing Before/After Projected Standings & Team Analysis for Add/Drops.
  • Finding Trades which shows your gains in Roto Standings or H2H Win%.
  • Find Free Agents which also shows your gains in Roto Standings or H2H Win%.

The goal is the create more accurate and consistent results between pages.  It also reduces the possibility of creating unrealistic results.  Feel free to let us know if other options/display values would be helpful.

Also, as mentioned, this is intended to replace some of the older pages, but we have also kept these available if you find their flexibility useful.


100% Agree (10 votes)
avatarfbbaddict (36 posts)  9/16 5:44 AM

My understanding is that (pls correct me if I'm wrong) Analysis Monster uses total game values, which I almost never use for H2H leagues, as it's impossible to nail "games played" accurately. It increases the possibiliy of creating unrealistic results imo. It'd be good to have an option to analyse with per game values. Again, I might be wrong, please correct me if that's the case.


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/16 6:00 AM

If you select "Assume 100% games", it's basically a per-game result since it assumes the same # of games for all players.


100% Agree (2 votes)
avatarfbbaddict (36 posts)  9/16 6:11 AM

Ah yes, it does the trick. Can we have an option to exclude IR from analysis completely? Also, saving the settings, so that we won't need to check "assume 100% games" each time. Cheers!


100% Agree (1 vote)
avatarken_admin (1534 posts)  9/16 6:15 AM

Thanks, and yes, we can save the setting.  Why would you want to exclude IR?


0% Agree (0 votes)
avatarfbbaddict (36 posts)  9/16 6:28 AM

In some cases, I want to see how my team or my opponents would do without their injured players in a specific period. Also, some injuries can be so tricky that teams won't provide a definite timeline. But nevermind, the old projected standings can be used for that. Thanks again Ken.


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/16 6:37 AM

For IR players, the use 100% will only use games after their projected to return date so they're weighted less.  Maybe that helps?


0% Agree (0 votes)
avatarfbbaddict (36 posts)  9/16 6:52 AM

Yeah it surely helps. I rarely need to check how it looks like when IR players are completely ignored, and as I said the old projected standings would help there.


0% Agree (0 votes)
avatardasein  (85 posts)  9/16 2:20 PM

What does apply game limits do, and when do you suggest using it?


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/16 2:44 PM

This is for Roto leagues that have Game Limits under League Settings.  It should always be applied, but we allow removing it to see what your complete team's games would look like.


0% Agree (0 votes)
avatardasein  (85 posts)  9/16 4:31 PM

Thanks. Which tool would be best for comparing teams that have an unequal number of players, on a per game basis? It doesn't seem like Analysis Monster is ideal for that.


100% Agree (1 vote)
avatarken_admin (1534 posts)  9/16 4:33 PM

Do you have a drafted league, I can look at?


0% Agree (0 votes)
avatarcarlos2898  (1 posts)  9/16 5:21 PM

The trade monster seems to be stuck on Totals. I try switching back to Per Game but doesn't switch from Totals. 


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/16 5:49 PM

We may need to add the Per Game value back, and we'll give that some thought.  I believe if you select "Assume 100% Games", that you'll pretty much be getting a Per Game value since all players get the same # of games.


0% Agree (0 votes)
avatardasein  (85 posts)  9/16 5:52 PM

@ken, for slow drafts that incorporate pick trading, or use auctions, it is common to have an inbalance in the number of players teams have at a given point during the draft. Methods that only use totals are not going to give useful results during a draft when a few teams have 10 players and a few have 7. You could look at my Heroes league as an example.

I think the best tool for comparing uneven teams currently is the old Projected Standings using Per Game, but I'm not sure. The usual reason I want to check this is to help decide which players to target during a draft.


100% Agree (1 vote)
avatarken_admin (1534 posts)  9/16 6:06 PM

I think you're right that Projected Standings is best during a live draft since Trade Monster, for now, assumes complete rosters.  We'll look at allowing the Per Game there too.


0% Agree (0 votes)
avatarJohnjclass101@gmail.com  (1 posts)  9/17 12:59 AM

Is the projected standings portion of analysis monster taking benched games into account? For example in a points league with no game limit, you could have two teams in a matchup with 39 total games each, but one team can only actually start 32 of them while because of schedule quirks, the other team can start 38 of 39. The head to head weekly feature doesn't take that into account, so I'm wondering if analysis monster does.


0% Agree (0 votes)
avatarKai-rie (54 posts)  9/17 2:47 AM

as soon as one of my guys is injured, it appears it just removes the player totally from the team analysis Is it possible to have a setting where it replaces the injured player with the next best replacement player for the number of games the injured player is expected out?


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/17 11:36 AM

@johnjclass - Analysis Monster is going to give you a general rankings with an assumed # of bench players.  It shows the % of games for bench players for the analysis.  For H2H Weekly, it should take actual games into account.  What are you seeing?


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/17 11:37 AM

@Kai-rie - For IR players, it will use that player's games and then take those games away from your bench players.  What are you seeing, and do you have an example league?


0% Agree (0 votes)
avatarterpoftroy  (4 posts)  9/19 10:13 PM

@ken My league is H2H points but with a 30 game per week limit. I am allowed to set lineups daily, but when I run the Analysis Monster with the daily lineup frequency setting, I just get an analysis assuming all projected games for the full roster, starters and bench. So the teams with full rosters are projected to do better than teams with less than full rosters (having a full roster is not a requirement in this league).

It gets a little bit better if I switch the lineup frequency setting to weekly. The analysis on this setting shows 820 games for all teams (which I assume is 10 active roster spots x 82 NBA games).

I think our season is 22 weeks, so the total games in the analysis should really be 30 games per week x 22 weeks = 660.

Does this make sense? Maybe the league is too nuanced for this tool; but effectively, the analysis should be of the best 660 games on the roster (starting with the best player's projected games, then the next best player's projected games, and so on until there are 660 projected games)

Appreciate all your work!


0% Agree (0 votes)
avatardasein  (85 posts)  9/20 5:57 AM

I'm just playing with AM now for a league that has drafted, albeit with uneven rosters. If I'm reading it right it assumes 100% games for the top X number of players on a roster and then splits the unused games equally among the remaining players. For a dynasty league, like this one, this is unhelpful because there are develpment players rostered that I would never start. I would think it better to make up the games not accounted for by the top X players, firstly by maxing out the X+1  best player, and when they are exhausted, the X+2 best player, and so on.

For example, this league has 10 starting spots so my 10 most highly rated players are assumed to start 100% of their projected games. My 11th best player is Monte Morris and my worst player is TyTy Washington. It seems that AM is assuming I will start both players in 25% of their games. In reality, Monte will start maybe 80% of his games and TyTy might not start ever this year.


100% Agree (2 votes)
avatarken_admin (1534 posts)  9/20 8:16 AM

@terpoftroy - Currently, we're not taking the weekly limit into account, but the way you describe it sounds like a good solution, and we'll implement that.

@dasein - That's probably a better solution, fill the bench with remaining games starting from best to worst players.


100% Agree (3 votes)
avatarsd21gr (2 posts)  9/21 11:10 PM

Hello guys,

been a member for a few years. I think the new update took the website a step back, kind of disappointing really. Firstly, you removed the very useful import all leagues button,its much more difficult to do now. Also, on the topic, the previous trade monster would let you add any player to any team, try 3team trades etc, it was very easy to recreate every scenario. Now its its only team a to b and free agents. Let me know if im missing something.

Thanks in advance


100% Agree (1 vote)
avatarken_admin (1534 posts)  9/22 4:19 AM

The "all rosters" link next to the league drop down will import all rosters so maybe that needs to be renamed?

For the trade monster, we did remove the manual adding of players to teams since it didn't seem too useful, but we can add that back.

Thanks for the feedback.


100% Agree (1 vote)
avatarsd21gr (2 posts)  9/22 9:52 AM

Thanks! Yes, I would rename to "import all rosters" or something if thats what it does. Also, if you add back manual adding, please do so ! :D


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/23 1:39 PM

Yes, we'll allow selecting between evenly using bench or use bench from best to worst until game limits are reached.


100% Agree (2 votes)
avatarpokatelli (2 posts)  9/26 8:33 PM

Perhaps I'm overlooking something but is there an option to adjust "% games to use for bench players" on the new analysis monster as there is on the old projected standings page? Currently i am only seeing analysis with this note...

100% Games for Bench Players
 
Using for a dynasty league where typically that value is set at 0% so that only 10 active players are calculated. Ideally would like to set this way on both the analysis monster and trade monster.
 
Thanks

0% Agree (0 votes)
avatarLuckykid  (6 posts)  9/27 1:21 AM

Hoping someone could help analyze my team 


100% Agree (1 vote)
avatarken_admin (1534 posts)  9/27 5:11 AM

@pokatelli  - Which of your leagues is this?  Is it roto, and if so, are there game limits? I'm just trying to see what other settings may be needed to address your setup.


0% Agree (0 votes)
avatarShpitzer8  (1 posts)  9/28 1:01 AM

Thanks, both are great tools. Nevertheless, I think there is a small bug in the calculation of the Analysis Monster: The standard deviation of the TO categories are regarded as (+) is "bad", (-) is "good". As a result, I think the H2H matchups are not calculated correctly. Thanks for your help


100% Agree (1 vote)
avatarKoksi (1 posts)  9/28 7:30 AM

I need help, I imported my league settings, and the ranking projections had lamelo being the 4th best player, that is due to my league settings concering how the points work, correct? 3pt made 3 etc.. 


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/28 8:19 AM

@Shpitzer8 - Do you have an example league showing this?  I think it may be coloring the value wrong, but the win % looks OK.


0% Agree (0 votes)
avatarpokatelli (2 posts)  9/28 1:15 PM

@ken_admin It's a CBS 8cat roto league with no game limits. Ideally would like to see this option in settings on the analysis monster, as it is in the old projected standings page...

Thanks


0% Agree (0 votes)
avatarken_admin (1534 posts)  9/28 1:19 PM

Are they weekly lineups or daily changes?


0% Agree (0 votes)