October 19, 2025
Why did I choose this as my topic, well I am in a league right now and I will see if charts do point out the obvious picks well or not. But for some guidance hopefully this presentation what will be perceived as professional for a NFL analysis:
We’ll model some positions stats (passing,rushing,receiving) as a function of points.
General Model:\[ Points = \beta_0 + \beta_1 \cdot Yards \]
R code:
lm(Points ~ Yards, data=fantasy)Purpose:
Outcomes:
Model For QBS:
\[ Points = \beta_0 + \beta_1 \cdot Passing \space Yards \]
qb_model <- lm(`Points (ESPN Scoring)` ~ `PASS YDS`, data = qbs) summary(qb_model)
Call:
lm(formula = `Points (ESPN Scoring)` ~ `PASS YDS`, data = qbs)
Residuals:
Min 1Q Median 3Q Max
-87.561 -32.801 -4.301 21.199 102.386
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 79.72409 41.52875 1.920 0.0636 .
`PASS YDS` 0.05339 0.01140 4.684 4.66e-05 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 49.59 on 33 degrees of freedom
Multiple R-squared: 0.3994, Adjusted R-squared: 0.3812
F-statistic: 21.94 on 1 and 33 DF, p-value: 4.661e-05
Model For Rbs: \[ Points = \beta_0 + \beta_1 \cdot Rushing \space Yards \]
rb_model <- lm(`Reg POINTS` ~ `RUSH YDS`, data = rbs) summary(rb_model)
Call:
lm(formula = `Reg POINTS` ~ `RUSH YDS`, data = rbs)
Residuals:
Min 1Q Median 3Q Max
-33.565 -9.540 0.000 8.298 52.242
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 41.128288 4.331297 9.496 2.87e-14 ***
`RUSH YDS` 0.147582 0.005417 27.244 < 2e-16 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 17.01 on 71 degrees of freedom
Multiple R-squared: 0.9127, Adjusted R-squared: 0.9115
F-statistic: 742.2 on 1 and 71 DF, p-value: < 2.2e-16
Model For WRs: \[ Points = \beta_0 + \beta_1 \cdot Receiving \space Yards \]
wr_model <- lm(`Points` ~ `REC YDS`, data = wrs) summary(wr_model)
Call:
lm(formula = Points ~ `REC YDS`, data = wrs)
Residuals:
Min 1Q Median 3Q Max
-32.595 -6.140 -1.575 6.360 27.618
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 2.662902 3.072300 0.867 0.388
`REC YDS` 0.144678 0.003251 44.507 <2e-16 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 9.898 on 93 degrees of freedom
Multiple R-squared: 0.9552, Adjusted R-squared: 0.9547
F-statistic: 1981 on 1 and 93 DF, p-value: < 2.2e-16
plot_ly(
qbs,
x = ~`PASS YDS`,
y = ~`PASS TD`,
z = ~`Points (ESPN Scoring)`,
type = "scatter3d",
mode = "markers",
text = ~paste(
"Player:", `First Name`, `Last Name`,
"<br>Yards:", `PASS YDS`,
"<br>TDs:", `PASS TD`,
"<br>Points:", `Points (ESPN Scoring)`
),
hoverinfo = "text",
marker = list(
color = ~`Points (ESPN Scoring)`,
colorscale = "Viridis",
size = 3
)
) %>%
layout(title = "QB Fantasy Points from Passing Yards and TDs")
At a glance we can do some takeaways for each of their on positions thats yards to matter and the graph can very understanding that more yards equal the more points. We also have to consider other dependent variables such as touchdowns, interceptions, fumbles. Then we can tell more of a story within the graphs.
For quaterbacks, maybe I shoudlve added passing touchdowns, and interceptions since those are the common stats in fantasy
For running backs, I think how I did it was fine maybe if i wanted to include receiving yards since that is also a common stat within a game as well.
For Receivers, this is usually the best way to determine because yards do matter but you also have to consider longest reception, reception + a catch, etc…etc.
All in all I think maybe a continuation after this I can come back and revise it to what I explained and see if any of the graphs change or stay the same. But also you have to consider alot of things as well. If a team plays well against zone focus defenses or man to man. Is there run blocking good? Is there pass protection good against the blitz, but thats more for sports betting.
