Category: Analytics

How do advanced metrics equate to strength of schedule?

Advanced metrics have not been kind to the Vikings this season, but how does it equate to strength of schedule?

You’ve heard it all before: the Minnesota Vikings are lucky and aren’t nearly as good as their 12-3 record would indicate. In fact, there are plenty of analysts, in both good and bad faith, that would tell you just that.

We’ve heard it all at this point. The astounding NFL record in one-score games at 11-0, the NFL record eight fourth-quarter comebacks and the “luck” they have had in such games, most notably the Josh Allen fumble. There is, however, more to projection than just talking about the underlying advanced metrics. There is a lot of context that needs to be taken into account when having such nuanced discussions.

One metric that a lot of people like to mention is DVOA from Football Outsiders. It’s a metric that tries to project forward based on how you have played thus far. One thing that is interesting is how the top teams in DVOA have had a weaker strength of schedule.

Now, DVOA as Akash mentions in the next tweet is a weighted metric that grades your performance based on who your competition is. What is really intriguing is that two of the top three are in the bottom three for strength of schedule while four and five have an SOS of 23rd and 27th.

How do the Vikings fare in comparison? They rank 25th in total DVOA at -9.4% with their SOS being at 14th at 1.1%.

What can we read from this? I think the biggest thing to read from this is that the Vikings, outside of their two massive duds, play to the level of their opponent. It’s why the Vikings have played so many one-score games this season and that explanation makes total sense.

That, and creating their own luck by outscoring opponents by 75 points after the third quarter.

What is HoopsHype’s All-World Draft Ranking?

The 2023 draft could be the first in history to feature two players from professional leagues in the Top 2 spots. Given how things have trended recently, it is surprising that it has taken so long for this to happen. While every NBA team has …

The 2023 draft could be the first in history to feature two players from professional leagues in the Top 2 spots. Given how things have trended recently, it is surprising that it has taken so long for this to happen. While every NBA team has international scouts covering various regions, it can be challenging to compare players from different competitions. So as a way to provide more clarity on how prospects are performing this season, we have developed the All-World Draft Rankings.

This metric tracks 50-plus competitions around the globe, ranking all draft-eligible players in professional leagues with no age limit. They range from the almost NBA-level G League and Euroleague to more humble domestic leagues in countries such as Ukraine and Georgia, as well as competitions in Asia, Latin America, Africa and Australia.

The All-World Draft Rankings are based on our Global Rating but factor in the strength of competition and date of birth, as well as other factors. For example, a 2004-born player in a strong league – such as Victor Wembanyama (France) or Scoot Henderson (G League) – will have a higher chance of making the ranking than a 2001-born player in a weaker tournament – such as Japanese sensation Yuki Kawamura.

It is important to note that these rankings are not intended to be a mock draft or a big board. They simply measure the performance of prospects during the 2022-23 season, without taking into account their NBA potential.

You can already check this new feature at the bottom of the main Global Rating page here.

Vikings overcame a massive win probability discrepancy vs. Colts

The Vikings continue to beat the odds

It’s no secret that the Minnesota Vikings had such a small win probability against the Indianapolis Colts at its worst that you couldn’t even give it a full percentage point.

It really shouldn’t be too much of a surprise either, as the Vikings were down 33-0 and were playing atrocious football. The chart itself looked like a plateau and peaked at a whopping 99.6% with 7:36 left in the third quarter.

It goes to show how rare a comeback like this is. Teams that led by 30+ points were 1,548-1 going into week 15 and the Vikings bucked the trend once again.

Eurobasket grades: How did NBA players perform?

It’s the NBA’s most boring month, but European basketball was at its best in September. The 2022 Eurobasket will be remembered as a fascinating and prediction-busting tournament, filled with NBA superstars and big upsets at every stage. Up to 35 NBA …

It’s the NBA’s most boring month, but European basketball was at its best in September. The 2022 Eurobasket will be remembered as a fascinating and prediction-busting tournament, filled with NBA superstars and big upsets at every stage.

Up to 35 NBA players represented their National Teams. This is how each one performed in arguably the most exciting Eurobasket of all time.

Eurobasket Global Rating: The best-performing players in the tournament

The 2022 Eurobasket could easily be the most star-studded in the tournament’s 87-year history. It includes recent winners of the most prestigious NBA accolades, such as MVP, Defensive Player of the Year and Rookie of the Year along with plenty of …

The 2022 Eurobasket could easily be the most star-studded in the tournament’s 87-year history. It includes recent winners of the most prestigious NBA accolades, such as MVP, Defensive Player of the Year and Rookie of the Year along with plenty of All-Star and All-NBA nods.

But are NBA superstars truly shining in the top European competition? Using Global Rating, we rank every player in the tournament so far. We will update the numbers every day until the final game on September 18.

Real Value: Which players are overpaid/underpaid in the NBA?

The offseason is the most stressful time for NBA front offices. Their decisions in the draft, the free agency and the trade market will not only impact their teams for years to come but their own careers as executives. To get in the mindset of …

The offseason is the most stressful time for NBA front offices. Their decisions in the draft, the free agency and the trade market will not only impact their teams for years to come but their own careers as executives. To get in the mindset of decision-makers around the league, we present Real Value, a tool designed to calculate how valuable each and every player in the league is according to their productivity on the court, age, potential and other intangibles. A full explanation of Real Value can be found here.

The current results are presented in two tables. The first one is the Real Value of every player under contract for 2022-23, including non-guaranteed and two-way players. The second one shows how that Real Value compares to their actual 2022-23 salaries to find out in a player is underpaid or overpaid.

How HoopsHype’s Real Value works

Valuing every player in the league, no matter if they are under contract or free agents, is one of the most important duties for NBA execs around the NBA. To bring that knowledge to you without having to ask your favorite GM, we introduce Real …

Valuing every player in the league, no matter if they are under contract or free agents, is one of the most important duties for NBA execs around the NBA. To bring that knowledge to you without having to ask your favorite GM, we introduce Real Value, a new statistical measure to calculate the market value of every NBA player.

Using Global Rating as the main performance metric, Real Value takes into account the productivity of each player from the last three seasons, giving extra value to the official NBA games played in the last 365 days. That includes regular season, play-in and playoffs. With that base, we calculate how that would translate to the “deserved” average salary per year.

Also factoring into Real Value: A player’s age, availability, defensive performance and off-court issues. For example, if a player gets admonished by his team due to detrimental conduct, it will be directly reflected in his value. Our NBA insider Michael Scotto will help us with direct insight from executives, coaches and scouts around the league to make Real Value as accurate as possible.

For some of the league superstars, the final number may reflect an unrealistic value, way over the maximum salary that a player can get as a free agent. But displaying that value is important to show how clearly they deserve max player consideration, and to put into context how team-friendly their contracts are.

For players whose Real Value falls under their minimum possible salary, the number is adjusted to that minimum. Those in that category are usually young players with little play time in the NBA and veterans who enjoy more of a mentorship role in their teams. Because they may need more minutes to showcase their full skills or due to the intangibles that are keeping relatively unproductive players as valuable full-time players in NBA rosters, their projected minimum is established as their Real Value floor.

The final number of every player will change almost every day for several reasons. On one hand, the algorithm takes into account games played in the last 365 days, so a new good game or a bad game from a year ago removed from the calculation will improve their Real Value, and the opposite. On the other, the money around the league is not infinite. The numbers will adjust considering the total money spent on salaries in the NBA. So if a franchise signs a player to a maximum contract adding $45 million to the 2022-23 pool, Real Values around the league will adapt accordingly.

As with every advanced metric, Real Value is meant to be used as a tool to simplify as objectively as possible an incredibly complex reality. At the end of the day, any player is as valuable as franchises decide for their own interests. Finding a consensus that’s as close to reality as possible is the goal here.

Is Celtics-Warriors Finals signaling the end of the super team era?

With the Golden State Warriors and the Boston Celtics in the NBA Finals, it is fascinating to watch how the two teams are very much alike in many ways. One of them: how their rosters were built. The stars in Boston and Golden State were drafted and …

With the Golden State Warriors and the Boston Celtics in the  NBA Finals, it is fascinating to watch how the two teams are very much alike in many ways.

One of them: how their rosters were built.

The stars in Boston and Golden State were drafted and cultivated from their respective teams and the same can be said about multiple role players.

For Boston, Jaylen Brown, Jayson Tatum, Robert Williams, Grant Williams, Marcus Smart, Payton Pritchard, Aaron Nesmith, and Sam Hauser have all only called the Garden home in their professional careers.

Stephen Curry, Klay Thompson, Draymond Green, Jonathan Kuminga, Kevon Looney, Moses Moody, Jordan Poole, James Wiseman, and Juan Toscano-Anderson all have been with the Warriors their entire careers.

The amount of home-grown talent is highly unusual in NBA Finals matchups.

But to what extent?

Well, per HoopsHype research, Warriors/Celtics players who have not played for any other team have combined for 69.8 percent of the minutes available at the 2022 Finals so far.

That would be the highest number since 1968.

Could this be the decline of the super team, or is this merely a fad?

Looking back at the past two seasons, the Milwaukee Bucks and Phoenix Suns of the 2021 NBA Finals came in at 35.97 percent, while the Los Angeles Lakers and the Miami Heat from 2020 were at 32.95 percent.

In the LeBron-heavy Finals of the 2010s, it never went above 58 percent.

However, it is important to remember James is well known for getting mercenary role players to help him win a title, and trading away all the homegrown talent from the teams he was on.

With the NBA being such a copycat league, it would not be surprising if more teams in the future use the approaches of the Warriors/Celtics to cultivate their own talent for the bulk of their rosters. Watching a team like the Lakers trade away all their young core only to see them implode this season could serve as a cautionary tale for franchises to be wary of.

HoopsHype’s Alberto de Roa contributed research to this report.

Analytics MVP 4.0: Who are the best players based on advanced analytics and impact metrics?

How do Joel Embiid and Nikola Jokic compare?

As part of a series at HoopsHype, we are examining who should win the NBA MVP award based on what we can learn from advanced analytics.

For this survey, each impact metric was included because it was considered among the most trustworthy by NBA executives when asked by HoopsHype during this past offseason.

The metrics pulled included Daily Plus-Minus (DPM), Estimated Plus-Minus (EPM), LEBRON (BBall-Index), RAPTOR (FiveThirtyEight), Player Efficiency Rating (Basketball-Reference), Box Plus-Minus (Basketball-Reference). We also added the alternative model of Box Plus-Minus from Backpicks.com as well as the newest impact metric, Daily-Updated Rating of Individual Performance (DRIP).

For the first time thus far, ESPN’s Real Plus-Minus (RPM) was also included.

Because all of these metrics (except PER) are graded on a per-100 possession scale, we adjusted for playing time by multiplying their impact contribution on each metric by the percentage of possible minutes they have played for their team so far this season.

As with the official vote, the top player received 10 points, the second received seven points, the third received five points, the fourth received three points, and the fifth player received one point. If a player finished outside of the Top 5, they didn’t receive any votes from that measurement.

Only players that made the Top 5 on least one of these nine metrics were included in our rankings below. Some of the most notable omissions include Kevin Durant, Rudy Gobert, Devin Booker, Jrue Holiday, Donovan Mitchell, Jimmy Butler, Fred VanVleet, James Harden, and Ja Morant.

To see who is performing the best based on HoopsHype’s Global Rating, click here. All stats are accurate as of March 27, 2022. 

Analytics MVP 3.0: Who are the best players based on advanced analytics and impact metrics?

We need to talk about Nikola Jokic.

As part of a new series at HoopsHype, we are examining who should win the NBA MVP award based on what we can learn from advanced analytics.

For this survey, each impact metric was included because it was considered among the most trustworthy by NBA executives when asked by HoopsHype during this past offseason.

The metrics pulled included Daily Plus-Minus (DPM), Estimated Plus-Minus (EPM), LEBRON (BBall-Index), RAPTOR (FiveThirtyEight), Regularized Adjusted Plus-Minus (RAPM), Player Efficiency Rating (Basketball-Reference), Box Plus-Minus (Basketball-Reference). We also added the model of Box Plus-Minus from Backpicks.com as well as the newest impact metric, Daily-Updated Rating of Individual Performance (DRIP).

ESPN’s Real Plus-Minus was not included because the data has not been published yet this season. However, if it’s publicly available by our next update, RPM will also be calculated.

Because all of these metrics (except PER) are graded on a per-100 possession scale, we adjusted for playing time by multiplying their impact contribution on each metric by the percentage of possible minutes they have played for their team so far this season.

As with the official vote, the top player received 10 points, the second received seven points, the third received five points, the fourth received three points, and the fifth player received one point. If a player finished outside of the Top 5, they didn’t receive any votes from that measurement.

Only players that made the Top 5 on least one of these nine metrics were included in our rankings below. Some of the most notable omissions include Rudy Gobert, Karl-Anthony Towns, Kevin Durant, Jimmy Butler, Jarrett Allen, Mike Conley, Donovan Mitchell, Devin Booker, and Ja Morant.

To see who is performing the best based on HoopsHype’s Global Rating, click here. All stats are accurate as of Feb. 17, 2022.