BuzzerBeater Forums

Here it is the pic again:
http://i59.tinypic.com/2cf9ova.jpg


Edit Perpete:
For this time, I'll edit the message, but deleting a message is really easier. If someone doesn't want his message deleted, he can just follow the rules and consider any other users has a human being not worthy of being insulted, no matter how harsh the level of that insult is.

Edit Lemonshine: I reposted it, so no need to amend the rest. I will only say that the person I replied to made extensive use of sarcasm and was very personal. Maybe I did go even further, but honestly it would be hard to call that a real insult.

Last edited by Lemonshine at 5/18/2015 8:06:05 AM

We can agree on that.

Also, as you can see sometimes there is a clear disconnection between the Official GS and the Real GS (extrapolated from DMI). For example note Week12 and Week14 for player 1 or, even more evidently, Week10 to Week14 for player 2 (especially in week 14 where the official GS is 7, but the DMI suggests a 9). I think it's possible that this inconsistency is caused by some adjustment been made to account for then number of games in a week (week 13 and 14 can have only 2 games). This is also speculation.

Here is my 2 guys on my main team with 9 stamina:
Player1..............DMI...Official GS...Est GS...Prev week mins...Change
Week5(purchase)...418900......7......7.09......87
Week6....................478800......7......7.40......78......0.31
Week7....................804900......8......8.27......94......0.87
Week8....................590300......8......7.79......63......-0.48
Week9...................1380400......9......8.93......63......1.14
Week10.................1380400......9......8.93......83......0.00
Week11.................1219400......8......8.79......79......-0.14
Week12...................737700......8......8.14......65......-0.65
Week13...................485400......7......7.43......84......-0.71
Week14...................758600......7......8.18......61......0.75

Week1(reset)........386300......7......6.87......52......-1.31
Week2...................477900......7......7.39......71......0.52
Week3...................813300......8......8.28......72......0.89

Player2..............DMI...Official GS...Est GS...Prev week mins...Change
Week5(purchase)...163500......6 6.44......67
Week6....................658200......8 8.92......62......2.48
Week7....................678200......9 8.96......85......0.04
Week8....................322400......8 8.00......65......-0.96
Week9....................677300......9 8.96......62......0.96
Week10...................681700......9 8.96......70......0.00
Week11...................680800......9 8.96......64......0.00
Week12...................412700......8 8.35......61......-0.61
Week13...................684400......8 8.97......55......0.62
Week14 683500......7 8.97......46......0.00

Week1(reset).........192000......7......6.96......32......-2.01
Week2....................358200......8......8.15......73......1.19
Week3....................278600......7......7.76......67......-0.39

This is a cool way to break it down. Thank you for sharing this.
So, let's see here, as far as the disconnect goes, I don't see it on player one in week 12. From my experience and based on forum discussions, 7 and 8 GS ratings have sub-levels, and 9 is a cap. If you're at 9, you are maxed out. This is all reflected in the DMI. In week 12 at 8 GS, it just looks like you're riding in the mid-range for your player at 8 GS.

Week 14 is in that funky place where at the end of the last week of the season all GS resets to 7, but DMI doesn't reset for several days while all the roster updates are made. I'm pretty sure that's what is happening here in both cases. Week 10 on player 2 looks like a regular 9 GS with a slight drop in skills/stamina since week 7 (you can see a slight increase in skills/stamina in week 10).

As far as GS changes, let me apply my probabilities to player 1 and see how they line up.
Week 6: GS stayed even at 7 with 78 mins - 44% chance of staying at 7 (not too crazy)
Week 7: +1 to 8 after 94 minutes - I have no data in this sample
Week 8: Stayed even at 8 with 63 mins - 45% chance of this, okay
Week 9: Jumped to 9 after 63 mins - 42% chance of this, seems okay
Week 10: Stayed at 9 after 83 mins - 36% chance of this, kind of lucky, but not too weird
Week 11: Dropped to 8 after 79 mins - 46% chance, so not unlikely
Week 12: Stayed at 8 with 65 mins - 48% chance of this, so seems good
Week 13: Dropped to 7 with 84 mins - 43% chance of this
Week 14: Looks like a pop to 8 (based on DMI) with 61 minutes - 63% chance, nicely done.

Last edited by Iguanadon Joe at 5/18/2015 10:50:46 AM

My week1 is the GS in the first week of competitive games. Week14 is the day before game-2 in the finals, so yeah you're right since the GS formally resets during the offseason calculations right after the finals, while both GS and DMI reset again on the following Thursday.

I had a mistake in the numbers on player2, now there are no inconsistencies. I have updated my post.

Last edited by Lemonshine at 5/18/2015 12:56:02 PM

Thanks everyone for the comments, gives me something to think about. Probably as Perpete said mostly I'm just unlucky. I don't think this is a case of not managing minutes well as someone suggested.

I'm constantly monitoring minutes. Usually I have much better results than I'm getting and I usually don't get the minutes to land as well as they have so far this season. I guess we're only 2 weeks in though so I should give it time.

Here's the log of my players so far:
Player Status W1 W2 GS-W2
Tetsu Takabe Core 67 71 Respectable
Stelios Tsouktos Core 69 72 Respectable
Plácido García Core 71 76 Respectable
Artúr Bánk Core 75 69 Respectable
Mirosław Serafin Core 82 71 Respectable
Antônio Brasil Core 66 76 Respectable
Joso Mustak Core 82 68 Respectable
Margiris Mašiotas Core 72 80 Respectable

Motoyuki Yoshimoto Fill-in 47 47 Strong
Ichiroh Oohashi Fill-in 45 45 Strong
Sho Katou Fill-in 47 45 Strong


I don't feel to bad about where they've landed.

I like the idea from Lemonshine to track the players to get a better feel of how they're being impacted. I track a lot of things in this game so I'll just add this on to the list. Lemonshine, would be cool if you or someone (buzzer-manager) could turn your data into a tool.



Last edited by Brutus at 5/18/2015 4:49:08 PM

Unfortunately in buzzer-manager you can only "easily" track the official GS and for players who are still playing for the same team.

That said I usually ignore the "official" number and use the formula whose coefficient were estimated by Josef Ka. This accounts for sublevels as sometimes you get a very good "respectable" or a very weak "strong" which are actually very similar (the DMI is similar without improvements in the player's skills). Unfortunately the formula uses old salary values so you also need to convert the player's skills into a salary using the old salary estimate. If I ever get the time, I'll simplify the calculation method and share it here (at the moment I have a macro running both the salary and the GS calculation).