Thursday, February 03, 2011
3:00 PM - 4:00 PM EST
The LIVE portion of this chat will begin on
Thursday, February 03, 2011 at 3:00 PM EST.
This is your chance to ask about the 2/3 release and any other HD issues.
Seble - Can you elaborate a bit on the change to increase the average ratings of new recruits? Particularly, what new recruits does this apply to? All of them, or just division 1, or what? Also, have the averages of the top 50 or top 100 players in the country changed? Thanks! Jeff (coach_billyg - Hall of Famer - 2:24 PM)
I should have provided more information about this change yesterday. Basically what I've done is bump up the averages for most individual ratings (which in turn raises the overall), while also lowering the standard deviation. The end result will be to close the gap between the average player and the extremes (both high and low). This should help address the concerns about a steep dropoff from the very elite players to the next level of DI recruits.
How much were the average ratings increased, and was this across the board for all levels? Were the percentage of high/med/low potential categories changed? (girt25 - Hall of Famer - 2:27 PM)
The average overall rating will only go up about 3 points or so with this change. It does affect all players because we generate recruits as an entire pool at one time. We don't generate them by division or rating range or anything like that. The average max ratings will not change (other than FT being lowered a bit), so you shouldn't see that much of a change in potentials.
Could you explain what the logic behind increasing the overall average ratings of recruits was? Seems to me all this accomplishes is making new recruits better than players currently in the game and introducing yet another period of adjustment because of this. (armst24 - Hall of Famer - 2:29 PM)
As I mentioned, this change should close the gap between elite and good players, which has been the biggest complaint I've heard about the recruit generation logic. I don't believe you'll have to adjust much if at all in your recruiting strategies.
Just curious of the logic/thinking behind lowering recruit FT shooting. Was the HD average for incoming freshmen above the RL average? (ryanderson - All-Star - 2:33 PM)
Looking at the overall numbers, FT% in HD is significantly higher than it is in real life college basketball. The problem was that the average FT grades for recruits was too high, so it's being lowered. In addition the average max rating for FT is also being lowered. So the end result is that players will have the same amount of potential in FT shooting, but the average grade will be lower than it is now.
Regarding D1 recruiting, Recruit Ranking Ave of Total Rating #1-#5 775 #9-11 715 #24-26 614 #48-50 581 #74-76 564 #98-100 555 #124-126 533 #149-151 525 #174-176 516 #198-200 508 The problem I see is there is the same difference #25 and #175, as #10 and #25. What is your view of this situation and how will the new recruits look? (reinsel - Hall of Famer - 2:48 PM)
This change should address that issue. In the test world where I generated new recruits the difference between #9-11 and #24-26 is about 39 pts and the difference between #24-26 and #174-176 is 154 pts. Of course the rankings don't go strictly by overall rating, so there will be some variance to these numbers from world to world and season to season.
If recruits have already been generated but recruiting hasn'e begun, will those recruits have ratings changes? Thanks. (hofhof - Hall of Famer - 2:50 PM)
No, only recruits generated after today will use the new logic. The first world to be affected is Tarkanian.
•Added a fix so that recruits won't be generated with very low start rating and average/high potential how is this getting fixed - will guys who would otherwise have single digit start ratings and high potential get a higher start rating or will they no longer get high potential or something else? (fd343ny - Hall of Famer - 2:57 PM)
Basically, if a player is created with average or high potential and has a low start rating, the start rating will now be bumped up a little bit. So his max rating doesn't change at all, he just gets a head start on meeting that max. You'll still see guys with lower start ratings have high potential, but you won't see guys with really really low start ratings have high potential. But it's important to know that we're not reducing the amount of potential, just bumping up the start rating.
Why won't recruits be generated with a very low start rating and high/average potential? This should not be a problem; the real issue was how slow the initial growth was -- why wasn't that the part that was addressed? (girt25 - Hall of Famer - 2:59 PM)
To change the way that player improvement works would be a major change requiring a good chunk of development and testing time. This change is basically a stopgap until I have time to redo the improvement logic.
why does this release have to be implemented mid season? why can't it be rolled out at rollover for each world? (dacj501 - Hall of Famer - 3:00 PM)
When the engine was rewritten last year in a new technology, we lost the ability to roll out versions on a world by world basis. That's something I want to get working again, but for now it's not possible.
Seble, so far I think all the changes have added a realism to HD....good job. When you say "Lowered 2-point and 3-point FG%", it means what? By how much? Im worried about the mid-season impact on current teams. (chewchad - Hall of Famer - 3:02 PM)
Overall FG% will drop about 2-3%. It was just too high across the board. This change will affect all players, so there shouldn't be an advantage or disadvantage for certain teams.
•Lowered 2-point and 3-point FG% How are you making this happen? I doubt that there is a simple setting in the engine for FG% - does this mean that DEF is being made more effective or that some factor of offensive skill (a function of ATH, SPD, LP, PERI and IQ + ???) is being reduced some? not asking for the algorithms, but can you give us a sense of HOW this effect will arise? (metsmax - Hall of Famer - 3:05 PM)
Actually this is a pretty straightforward change. Basically a player's shooting ability for a particular range is translated to a base FG%. I just modified that translated to lower FG% across the board. This is done before factoring in all the other variables (e.g. passing of teammates, defense, etc.) There was no change made to the impact of defense.
Re: the assists change, I'm guessing this is a cosmetic change -- the engine simply handing out more assists, rather than a material change in the engine? (girt25 - Hall of Famer - 3:07 PM)
Yes, it's a cosmetic change. I was pretty surprised by how low the assist numbers were compared to real life.
Why are assists locked in so heavily by position? Shouldn't it be more related to their ratings. (I.E. right now you could play a sf at pf and his assists would go down, and play the same player at sg and his assists would go up. In real life that's not the case.) (girt25 - Hall of Famer - 3:08 PM)
Assists aren't that heavily locked into position. A good passing big man will still rack up some assists, just not as many as a PG with an equal passing rating.
When you say there is a small reduction to fatigue rate; does this scale for all players or is it 'fixed'? For example would a player with 50 stamina get the same exact minute increase by the reduced reduction as an 80 stamina player, or would that be a reduced fatigue rate? In other words, does a high stamina player benefit more just from this change than a low stamina player? (charleym14 - Hall of Famer - 3:09 PM)
The change to fatigue is to the base rate of fatigue, before stamina is factored in. So it should affect players about the same amount.
" Small reduction to fatigue rate" - Could you explain why this was necessary and how small a change. Many people are afraid that this will bring back the dominance of the FCP. (lmschwarz - Hall of Famer - 3:12 PM)
I looked over data related to the minutes per game in HD vs. real life, focusing on starters. The HD number is significantly lower than real life. I felt the fatigue rate was too high even before the new engine was released last year, but kept it based on feedback during user testing. This change does affect all players, so I don't see it making the press more effective.
By increasing steals, have you just increased the rate based on the existing factors, or have you reweighted or reworked which factors contribute to this higher steal rate? Should I still expect speed, defense rating, and defense IQ to play the biggest factor here? (charleym14 - Hall of Famer - 3:15 PM)
I've increased the base rate, so there's no change to the various factors that can increase or decrease that rate.
The new release says that it "reduces FT grades" -- does this mean the percentages reflected by the letter grades are changing or simply that incoming players won't shoot FT's as well? (i.e. -- will someone who is a 'B' FT shooter shoot the same percentage post-release as they did pre-release, or do I need to reconfigure mentally what it means to be a B now) (rednu - Hall of Famer - 3:18 PM)
Nothing is changing in how the engine interprets a player's FT grade. The new guys will just have lower grades on average.
Seble, congrats on the new release, looks like you have addressed some real issues here. Can you confirm that the highest priority for the next release will be to fix to architecture so it in possible to stagger future releases so they become effective when the world changes over and not in the middle of the season? (_hannibal_ - Veteran - 3:20 PM)
It's definitely high on the list. It's obviously not my preference to release a new engine version in the middle of a season, but I felt this change was not drastic enough to delay it indefinitely.
When will the much anticipated jobs process change be put into place? What will the actual change be? (stinenavy - Hall of Famer - 3:22 PM)
Still working on some things for the job process, and it will probably be the next significant release. I know that the logic for loyalty will be changing some, and most likely the way that success is calculated will change to incorporate more of a coach's history.
Do you have plans to review the cost structure of FSS? Thanks. (Rails - Hall of Famer - 3:23 PM)
Most likely there will be some kind of change made to FSS, but I'm not sure what that will be yet.
Have you noticed any data that shows that zone defense is tough to win with? Seems that Press and M2M win the lion's share of NT titles. (stinenavy - Hall of Famer - 3:24 PM)
I have noticed that zone teams have a lower overall winning percentage. I'm not sure if that's a product of the way the engine handles zone or if the teams running a zone just aren't as good. It's something I'll look at more closely.
So that would mean that we should not see more turnovers than before, just a higher percentage of turnovers will be steals as compared to before? (hughesjr - Hall of Famer - 3:25 PM)
Yes, turnover rate should remain the same, just more of those turnovers will result in steals.
Not sure how having the fatigue base rate adjusted helps prevent the FCP from becoming dominant again. The problem with FCP was adjusted by making player's ability drop off as they fatigued. If fatigue is not as big an issue now (even if it's not for the opponent either), the abilities won't drop off and FCP stays effective longer, i.e., no tired player penalty in the FCP. (dukenilnil - Hall of Famer - 3:27 PM)
This isn't a drastic change, so players will still get tired running a press. Even without considering fatigue, there are negatives to running a press that balance out the positives.
I feel REB still needs work. THe previous engine treated REB as the most important factor by FAR. Now i see guys with 50-70 REB holding their own or beating out guys with the 80-99 REB. I have a SG with a 1 REB who has pulled down 5 boards twice this year and 2+ several other times. Can you explain the logic behind how the new engine treats/awards rebounds compared to the old. Many of us feel the REB rating itself is nowhere near as important as it used to be. (jjboogie - Hall of Famer - 3:32 PM)
I get quite a bit of feedback on rebounding and often the problem is that a coach isn't looking at the whole picture. A lot of times a coach will look at only the PF and C on each team and base their conclusions on that. Or they'll ignore the fact that a great rebounder only plays 18 minutes per game. Or they'll look only at the total numbers in a game, without considering how many rebound opportunities were there for each team. I'm actually very confident that rebounding is working fine. Contrary to popular belief, the REB rating is by far the most important in determining a player's ability in rebounding.
Based on user statistics, the population of worlds is on a serious decline. What is on the horizon to help rejuvenate some of the worlds? (mlatsko1 - Hall of Famer - 3:34 PM)
Well, from my perspective I can only keep working to improve the game. Marketing and promotion are not my area. The decline has happened across all of our games to varying degrees, so it's not isolated to HD. We are very aware of it though.
Do you have any idea what the affect on d2 and d3 recruits will look like? At first it sounded like everyone was getting better, but now it is unclear. Do you think d2 recruits will roughly get better, worse, or stay the same (counting the bottom half of d1 recruits and d2 recruits)? What about d3 recruits (counting most of d2 recruits as d3 recruits)? (coach_billyg - Hall of Famer - 3:35 PM)
They will get slightly better on average. Remember that this change is only impacting start ratings, not max ratings.
I think the problem with the Starter's minutes is that coaches play too much "Fairly Fresh", which results is guys getting lower minutes than real life. I think we should be comparing minutes at "Tired" to real life and not Minutes at "Fairly Fresh". (hughesjr - Hall of Famer - 3:37 PM)
I'm sure that's true to some degree. Even with this change to fatigue, the HD MPG for starters will still be lower than real life. I don't agree that "Tired" players in HD should be compared to a normal real life player. A "Tired" HD player is experiencing a pretty significant dropoff in performance.
thank you for the response to REB - if you compare rebounding margins in RL to HD there is a huge discrepancy. top teams in RL average 13-14 in REBmargin. top teams in HD right now are 7.5 to 8 REBmargin. There is no elite rebounding team in HD anymore. If it's not an individual REB rating then the way teams get rebounds is off. (jjboogie - Hall of Famer - 3:38 PM)
I haven't looked at margins, but if that's the case, then the issue would be in recruit generation (i.e. too may good rebounders), not in the engine.
Any chance that prestige will be reviewed? If you are outside of the Big 6 it is a huge struggle to move up, but have an off season and you get a major thumping. I would like the chance to occasionally recruit outside of my backyard. Thanks. (mjwilson - Hall of Famer - 3:39 PM)
Yes, it's an area that will be reviewed at some point.
Seble, Players with 20 points of improvement, before getting the capped email, appear as medium potential in the recruiting search display, but as high potential in the assistant coach email. Is there any plans to fix this? Also, any plans to expand the medium potential category to sub potentials, like the highs and lows? Thanks, Jeff (coach_billyg - Hall of Famer - 3:40 PM)
I'm not aware of this happening. It is possible that a player is recruited as high potential and by the time the Player Thoughts email comes he is at average. That can happen if the player is right on the edge of the two ranges and improves enough to bump him to the next range.
When you say that you increased assists significantly, what exactly does this mean? Could we now see great passers frequently having double-digit assist games? (ryanderson - All-Star - 3:43 PM)
The DI assist leaders before this release were usually around 4 per game. After this change they'll be around 6 per game, possibly higher. The change will also lead to more high-assist individual games.
again, can you explain how this engine treats REB differently from the old? Or give us some idea what the basic needs are? My teams are quite simply not rebounding as well as they used to, and my composition has not changed a ton. If anything, I tend to care more about ATH than I used to. (wronoj - Hall of Famer - 3:53 PM)
The old engine basically determined the odds of offense or defense getting the rebound, then assigned the rebound to a particular player on whichever team got it. The current engine gives each of the 10 players on the floor a percentage chance to get the rebound based on ratings/etc. Then the rebound is assigned based on those individual odds instead of a formula for team odds. The formula for a player's ability in rebounding is fairly similar between the two engines.
You mentioned in the release *Increased steals (does not increase turnovers, just percentage of turnovers that are steals). If a team is FCP or HCP wouldn't a teams POT increase if more of the TO's are Steals now? Double Dribble, traveling etc. doesn't give a team FB scoring oppurtunity like a steal will do in RL. It wouldn't make sense if POT% wouldn't increase if steals are increased. Could you elaborate on the logic change? thx for the hard work! (bigteniowa - Hall of Famer - 3:54 PM)
Not a major increase, but a little bit.
If we remove the press as a defensive set and make it a situational option, then Zone v. Man can be balanced easier; fatigue can be more easily adjusted and MPG will be more likely to reflect real life. So when are you removing the press again? (mlatsko1 - Hall of Famer - 3:56 PM)
Lol, good question. While I'd like to change the way that pressing works in the game, it would be a very major change and would impact a lot of different areas. With our resources limited we have to prioritize things and look at time required vs. benefit. Right now there are other things that will be more beneficial than changing the press.
When you looked at the rl minutes of starters, in addition to concluding that minutes played were higher, did you also take note of the range of minutes played being pretty narrow? Doesn't that suggest that individual stamina levels are similar between rl college basketball players? (Rails - Hall of Famer - 4:00 PM)
Actually I don't see that being the case at all. In 2009-10, of players that played more than 20 games and started at least 80% of them, MPG ranges from 39.1 to 11.9 with an average around 29.
Does offensive or defensive set have any effect on the players chance to get a rebound? (mlatsko1 - Hall of Famer - 4:02 PM)
Defensive set does. Zone and Press will give up more offensive rebounds than Man-to-Man, all else being equal.
I know there are a lot of guys interested in upping the participation in HD to enhance the experience for all. There was one thread in particular with a number of ideas. Can you provide some contact information for the marketing and promotion departments so that people can get their ideas/concerns to the proper place? (_hannibal_ - Veteran - 4:03 PM)
I'm more than happy to pass along ideas for marketing/promotion to the appropriate people. Unfortunately coming up with ideas isn't the biggest obstacle. It usually comes down to money or time.
Mark Zuckerberg created Facebook in about a week (his words). Why do developments here take so long? j/k. Thanks for the good work on improving the game. (dukenilnil - Hall of Famer - 4:04 PM)
Mr. Zuckerberg is full of it.
While you are changing the starting points of ratings for new recruits, are you doing anything to handle recruit location at all? The reason I ask is, for certain areas, there always seems to be a half dozen studs to pick from, while others might only get on good player in their backyard every 3 or 4 seasons. I am talking about schools or states that have major populated cities to draw from that have very few top shelf player to recruit. (elmossle - Hall of Famer - 4:04 PM)
No change to geographical distribution in this release.
What about some of the other small but much requested improvements, such has players appearing on your roster as soon as they've signed and being able to put in the recruiting actions for the first cycle in advance (to name two of many)? (girt25 - Hall of Famer - 4:06 PM)
Moving signed recruits early is a much larger and more difficult change than it seems. Not sure there's enough benefit to invest the work. Extending the first recruiting cycle is a much more feasible change.
Fatigue vs Performance changes. Some lower Stamina players(75-85) seem to effected but not others. I still see players with 80 Stamina, averaging 20 a game playing 35+ min with 60% FG % in D3. Has this been addressed with this release or future releases? (spottratz - All-Star - 4:08 PM)
If a player is really good then they still may be effective despite some fatigue.
the changes to assists and steals are mostly cosmetic in that the end game results should be about the same, but more of the turnovers will now be steals instead of say passes out of bounds or dribbling errors. Does the offensive player's bh/pas have any effect on this? (dacj501 - Hall of Famer - 4:10 PM)
The offensive player's ability impacts whether a turnover occurs of course, but whether it's a steal is mostly determined by defensive factors.
Speaking of Fatigue settings, I'm a vet and I've never seen a clear answer to this. What does Foul Trouble Only mean? Does it mean I set my starter to FTO, then he will only come out if he's in foul trouble, or does it mean if I set one of my 325 rated D1 walkons to FTO, that he'll only come in when everyone else is in foul trouble? (mlatsko1 - Hall of Famer - 4:12 PM)
It means that the player will only exit the game when he's in foul trouble.
Still a lot of questions I couldn't get to, but hopefully this chat helps give some insight into what changed in today's release and why. ( Moderator - 4:15 PM)
Thanks for taking the time to participate.