Deep Machine Learning, what it thinks of Dak, and the Cowboys

cowboyec · Feb 22, 2020

zerofill said:
Unlike some people that talk I have built an A.I. setup... basically machine deep learning.

I have been drowned in work lately... I am a programmer,
who works with both normal programming, and A.I. Neural learning networks/Deep learning.

If you want to see what GPT-2 thinks of the Dallas Cowboys if they keep the
same coaches and the same players... with the upcoming team schedule next season,
and a random one, for the next seasons, this is what it came up with for the next
5 years.

7-9
8-8
8-8
11-5
12-4

NOW.. Granted this is not going to work out, because we have changed coaches,
so technically, I have very little data to put in, because I have to go by
weird stuff, that I add from Green bay, that doesn't really fit here.

But, as far as Dak goes, I have input every game he had this past season.

The A.I. puts him at 12-13 as far as QBs go in ranking.

O.k... @PAPPYDOG, this is probably higher than we put him lol.

Funny thing is, it actually puts Mahomes as the top QB until week 8, and
then puts Matt Ryan higher... I haven't investigated that yet.

It also thinks Baker Mayfield will pass for 5200 yards next season, so you never know.

Granted, machine learning is still in it's infancy, and I am trying to get it
to start reading, information about the cowboys and other teams that it finds on
the internet.

But, outside of normal work, I am working on a horse racing one first lol.

(So far, testing I spent $144 and won $378)

Anyone interested in that, let me know via PM. Not sure how many people are actually
interested in Machine Deep Learning stuff lol.

If you do PM me, just know that it may take me a couple days right now to get back
with you,

I will also, give the Dallas github repo to you, if you are a programmer.
Just also, PM me and let me know if you are a programmr. Because I will
just send it to you, versus bumping the post every time.

I figure probably a month or 2 months from now... no later than the end of
June, I will let everyone know what we came out as.

And then I will redo it again, right before regular season, using pre-season
data...

you back engineered this technology at area 51 using recovered alien technology from a crashed ufo didn't you....i know you can't answer.

Irvin88_4life · Feb 22, 2020

jaythecowboy said:
So we only needed to give Garrett 4 more years

He still wouldn't get out the divisional round. Garrett has been 13-3, better than 12-3, and still lost in the divisional.

RodeoJake · Feb 22, 2020

zerofill said:
Unlike some people that talk I have built an A.I. setup... basically machine deep learning.

I have been drowned in work lately... I am a programmer,
who works with both normal programming, and A.I. Neural learning networks/Deep learning.

If you want to see what GPT-2 thinks of the Dallas Cowboys if they keep the
same coaches and the same players... with the upcoming team schedule next season,
and a random one, for the next seasons, this is what it came up with for the next
5 years.

7-9
8-8
8-8
11-5
12-4

NOW.. Granted this is not going to work out, because we have changed coaches,
so technically, I have very little data to put in, because I have to go by
weird stuff, that I add from Green bay, that doesn't really fit here.

But, as far as Dak goes, I have input every game he had this past season.

The A.I. puts him at 12-13 as far as QBs go in ranking.

O.k... @PAPPYDOG, this is probably higher than we put him lol.

Funny thing is, it actually puts Mahomes as the top QB until week 8, and
then puts Matt Ryan higher... I haven't investigated that yet.

It also thinks Baker Mayfield will pass for 5200 yards next season, so you never know.

Granted, machine learning is still in it's infancy, and I am trying to get it
to start reading, information about the cowboys and other teams that it finds on
the internet.

But, outside of normal work, I am working on a horse racing one first lol.

(So far, testing I spent $144 and won $378)

Anyone interested in that, let me know via PM. Not sure how many people are actually
interested in Machine Deep Learning stuff lol.

If you do PM me, just know that it may take me a couple days right now to get back
with you,

I will also, give the Dallas github repo to you, if you are a programmer.
Just also, PM me and let me know if you are a programmr. Because I will
just send it to you, versus bumping the post every time.

I figure probably a month or 2 months from now... no later than the end of
June, I will let everyone know what we came out as.

And then I will redo it again, right before regular season, using pre-season
data...

I can appreciate your efforts. 30 some odd years ago I wrote sub routines in machine language that examined billing errors and fixed them. They also examined free form name and address fields and split them into separate fields. I was working on sub routines that would be called in the event of a system abend. They would investigate the problem, bypass the account, spit out a report, and allow the system to continue without interruption. I was told those weren't worth the man hours and drop my little project.

Rockport · Feb 22, 2020

RodeoJake said:
I can appreciate your efforts. 30 some odd years ago I wrote sub routines in machine language that examined billing errors and fixed them. They also examined free form name and address fields and split them into separate fields. I was working on sub routines that would be called in the event of a system abend. They would investigate the problem, bypass the account, spit out a report, and allow the system to continue without interruption. I was told those weren't worth the man hours and drop my little project.

Sounds like you’re talking about the old IBM mainframes running OS 360 or OS 390 which uses machine language at it’s core like Unix systems use C. Machine learning is totally different. Machine learning combines computer science and statistics to create statistical models, which are then used to make predictions about the world or to infer patterns in your data. The most common language used in machine learning is Python but several others are used as well.

Adreme · Feb 22, 2020

I am curious about how the machine was fed information as well as how the machine learning algorithm would be able to predict the quality of drafts for the next few years given that we do not know the quality of those positions.

Just a few questions, as someone who is at the moment not that skilled in making machine learning algorithms, but understands how they work pretty well, was the algorithm tested using previous years data to build the current pool of information and was it able to correctly guess the things that had already happened? Or to reword that sentence if I said it badly, if you supplied it with information from say prior to 2018 was it able to predict the results of the 2018 season?

RodeoJake · Feb 22, 2020

Rockport said:
Sounds like you’re talking about the old IBM mainframes running OS 360 or OS 390 which uses machine language at it’s core like Unix systems use C. Machine learning is totally different. Machine learning combines computer science and statistics to create statistical models, which are then used to make predictions about the world or to infer patterns in your data. The most common language used in machine learning is Python but several others are used as well.

Old mainframe technology using assembler language. Lots of bit manipulation in registers, translate tables and n immediate commands. The actual code for those sub routines was extensive and difficult to follow.

Rockport · Feb 22, 2020

RodeoJake said:
Old mainframe technology using assembler language. Lots of bit manipulation in registers, translate tables and n immediate commands. The actual code for those sub routines was extensive and difficult to follow.

Yeah I used to debug those abend dump print outs back in the day.

RodeoJake · Feb 22, 2020

Rockport said:
Yeah I used to debug those abend dump files back in the day.

Me too. It was always 2 or 3 in the morning. That was my motivation for creating an abend subroutine. Honestly, I don't really think it was the man hours that scared the big shots. I believe they were intimidated by something they were unable to grasp. They couldn't grasp the intricacies of assembler and really hated that I used it.

HungryLion · Feb 22, 2020

I’m going to start PM’ing you before I place bets on games.

garyo1954 · Feb 23, 2020

Yes, now that I've researched a bit more I'm interested in the program.
The amount of historical data seems to be more key than the number of variables used, although team data could be weighted to offset a new system and coaching staff.

Depending on the number of variables the data file could be enormous headache to compile and maintain too. I sat down last night and popped off 30+ factors that affect any game.

So yes! I am interested in seeing what you have.

Thanks!

ShiningStar · Feb 23, 2020

Good start, keep us posted.

PAPPYDOG · Feb 23, 2020

ll

zerofill said:
Unlike some people that talk I have built an A.I. setup... basically machine deep learning.

I have been drowned in work lately... I am a programmer,
who works with both normal programming, and A.I. Neural learning networks/Deep learning.

If you want to see what GPT-2 thinks of the Dallas Cowboys if they keep the
same coaches and the same players... with the upcoming team schedule next season,
and a random one, for the next seasons, this is what it came up with for the next
5 years.

7-9
8-8
8-8
11-5
12-4

NOW.. Granted this is not going to work out, because we have changed coaches,
so technically, I have very little data to put in, because I have to go by
weird stuff, that I add from Green bay, that doesn't really fit here.

But, as far as Dak goes, I have input every game he had this past season.

The A.I. puts him at 12-13 as far as QBs go in ranking.

O.k... @PAPPYDOG, this is probably higher than we put him lol.

Funny thing is, it actually puts Mahomes as the top QB until week 8, and
then puts Matt Ryan higher... I haven't investigated that yet.

It also thinks Baker Mayfield will pass for 5200 yards next season, so you never know.

Granted, machine learning is still in it's infancy, and I am trying to get it
to start reading, information about the cowboys and other teams that it finds on
the internet.

But, outside of normal work, I am working on a horse racing one first lol.

(So far, testing I spent $144 and won $378)

Anyone interested in that, let me know via PM. Not sure how many people are actually
interested in Machine Deep Learning stuff lol.

If you do PM me, just know that it may take me a couple days right now to get back
with you,

I will also, give the Dallas github repo to you, if you are a programmer.
Just also, PM me and let me know if you are a programmr. Because I will
just send it to you, versus bumping the post every time.

I figure probably a month or 2 months from now... no later than the end of
June, I will let everyone know what we came out as.

And then I will redo it again, right before regular season, using pre-season
data...

All respects as you are a member who only adds to the quality of this site!
My beef with Dak is that I firmly believe he does not have the God-given talents to play as a starter in the NFL.
He also needs mucho cap bucks on the offensive( Needs to be carried) and add to that he wants a top 5 contract for basic stats which are against teams and time which are Garbage!
If Jerry Jones gives this young man a mega long term deal it will be his biggest mistake since he let Jimmy go.
Personally I believe they're not SOLD on Dak so all is game come this spring.

Corso · Feb 23, 2020

PAPPYDOG said:
ll

All respects as you are a member who only adds to the quality of this site!
My beef with Dak is that I firmly believe he does not have the God-given talents to play as a starter in the NFL.
He also needs mucho cap bucks on the offensive( Needs to be carried) and add to that he wants a top 5 contract for basic stats which are against teams and time which are Garbage!
If Jerry Jones gives this young man a mega long term deal it will be his biggest mistake since he let Jimmy go.
Personally I believe they're not SOLD on Dak so all is game come this spring.

All is game. Yes.
Highly interesting, how this is going.

pansophy · Feb 23, 2020

zerofill said:
Unlike some people that talk I have built an A.I. setup... basically machine deep learning.

I have been drowned in work lately... I am a programmer,
who works with both normal programming, and A.I. Neural learning networks/Deep learning.

If you want to see what GPT-2 thinks of the Dallas Cowboys if they keep the
same coaches and the same players... with the upcoming team schedule next season,
and a random one, for the next seasons, this is what it came up with for the next
5 years.

7-9
8-8
8-8
11-5
12-4

NOW.. Granted this is not going to work out, because we have changed coaches,
so technically, I have very little data to put in, because I have to go by
weird stuff, that I add from Green bay, that doesn't really fit here.

But, as far as Dak goes, I have input every game he had this past season.

The A.I. puts him at 12-13 as far as QBs go in ranking.

O.k... @PAPPYDOG, this is probably higher than we put him lol.

Funny thing is, it actually puts Mahomes as the top QB until week 8, and
then puts Matt Ryan higher... I haven't investigated that yet.

It also thinks Baker Mayfield will pass for 5200 yards next season, so you never know.

Granted, machine learning is still in it's infancy, and I am trying to get it
to start reading, information about the cowboys and other teams that it finds on
the internet.

But, outside of normal work, I am working on a horse racing one first lol.

(So far, testing I spent $144 and won $378)

Anyone interested in that, let me know via PM. Not sure how many people are actually
interested in Machine Deep Learning stuff lol.

If you do PM me, just know that it may take me a couple days right now to get back
with you,

I will also, give the Dallas github repo to you, if you are a programmer.
Just also, PM me and let me know if you are a programmr. Because I will
just send it to you, versus bumping the post every time.

I figure probably a month or 2 months from now... no later than the end of
June, I will let everyone know what we came out as.

And then I will redo it again, right before regular season, using pre-season
data...

Challenging problem to model. Have you considered doing a supervised approach using 2017 and 2018 to predict 2019? Are able to see what features are being weighted heavily by the model?

Seven · Feb 23, 2020

PAPPYDOG said:
ll

All respects as you are a member who only adds to the quality of this site!
My beef with Dak is that I firmly believe he does not have the God-given talents to play as a starter in the NFL.
He also needs mucho cap bucks on the offensive( Needs to be carried) and add to that he wants a top 5 contract for basic stats which are against teams and time which are Garbage!
If Jerry Jones gives this young man a mega long term deal it will be his biggest mistake since he let Jimmy go.
Personally I believe they're not SOLD on Dak so all is game come this spring.

LOL!! SERIOUSLY .........in one post we went from leading edge technology to cow tipping 101. Regurgitated.........great post, Elroy.

FTWayne · Feb 23, 2020

Rockport said:
AI, machine learning and modeling are only as good as the data going in. Trash in, trash out.

hating troll, move alone.

Aviano90 · Feb 23, 2020

So he is on the same trajectory as Romo was?

JoeKing · Feb 23, 2020

zerofill said:
Unlike some people that talk I have built an A.I. setup... basically machine deep learning.

I have been drowned in work lately... I am a programmer,
who works with both normal programming, and A.I. Neural learning networks/Deep learning.

If you want to see what GPT-2 thinks of the Dallas Cowboys if they keep the
same coaches and the same players... with the upcoming team schedule next season,
and a random one, for the next seasons, this is what it came up with for the next
5 years.

7-9
8-8
8-8
11-5
12-4

NOW.. Granted this is not going to work out, because we have changed coaches,
so technically, I have very little data to put in, because I have to go by
weird stuff, that I add from Green bay, that doesn't really fit here.

But, as far as Dak goes, I have input every game he had this past season.

The A.I. puts him at 12-13 as far as QBs go in ranking.

O.k... @PAPPYDOG, this is probably higher than we put him lol.

Funny thing is, it actually puts Mahomes as the top QB until week 8, and
then puts Matt Ryan higher... I haven't investigated that yet.

It also thinks Baker Mayfield will pass for 5200 yards next season, so you never know.

Granted, machine learning is still in it's infancy, and I am trying to get it
to start reading, information about the cowboys and other teams that it finds on
the internet.

But, outside of normal work, I am working on a horse racing one first lol.

(So far, testing I spent $144 and won $378)

Anyone interested in that, let me know via PM. Not sure how many people are actually
interested in Machine Deep Learning stuff lol.

If you do PM me, just know that it may take me a couple days right now to get back
with you,

I will also, give the Dallas github repo to you, if you are a programmer.
Just also, PM me and let me know if you are a programmr. Because I will
just send it to you, versus bumping the post every time.

I figure probably a month or 2 months from now... no later than the end of
June, I will let everyone know what we came out as.

And then I will redo it again, right before regular season, using pre-season
data...

Don't care. It's still just an opinion.

Ranched · Feb 23, 2020

Redball Express said:
do a girl that striptease for you using my cellphone keyboard and my index finger to point where to strip.?

GMO415 · Feb 23, 2020

MM is a huge unknown, but I'm certain we'll be better than what Garrett could do

Deep Machine Learning, what it thinks of Dak, and the Cowboys

Well-Known Member

Well-Known Member

Well-Known Member

AmberBeer

Well-Known Member

Well-Known Member

AmberBeer

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

There are no Dak haters just Cowboy lovers!!!

Offseason mode... sleepy time

Well-Known Member

Messenger to the football Gods

Well-Known Member

Go Seahawks!!!

Diehard

"We Are Penn State"

Well-Known Member