A federal sentence can feel like it comes out of a black box. In reality, it is built through a defined process, and almost every step of that process can be contested.
The system uses the U.S. Sentencing Guidelines, a point-based framework that produces a recommended range of months in prison. The important part — the part that many people miss — is that the range is a starting point, not the final word.
We defend clients through federal sentencing, and our team includes a former Assistant U.S. Attorney who worked these cases from the government’s side. That experience shapes how we approach sentencing: not as a formality after a plea or verdict, but as a stage where the outcome is still very much in play.
This page walks through how the calculation actually works, step by step, and shows where an experienced defense can change the number.
If you or someone you love is facing federal sentencing, the calculation is not fixed yet. Contact our office to talk about what can still be challenged.
The Guidelines Are Advisory, Not Mandatory
Before the math, one point changes everything that follows. For years, judges were required to sentence within the guideline range. That is no longer the law.
The Supreme Court made the Sentencing Guidelines advisory. A federal judge still has to calculate the range and consider it, but the judge is not bound by it and can impose a sentence above or, more importantly for the defense, below it.
That single fact is why sentencing is a real contest and not just arithmetic. The calculation sets the reference point, and then a separate set of arguments can pull the sentence away from it.
Step One: The Base Offense Level
Every federal offense is assigned a base offense level, found in Chapter 2 of the Guidelines Manual. Offense levels run from 1, for the least serious conduct, up to 43 for the most serious.
The base level depends on the crime. Many fraud offenses start at level 7, for example, while more serious crimes start much higher.
This number is only the foundation. From here, the level moves up or down based on the specific facts of the case.
Step Two: Specific Offense Characteristics
Next, the guideline for your offense lists specific offense characteristics — facts that add levels because they make the conduct more serious. These stack on top of the base level.
In a fraud case, the dollar amount of loss is usually the biggest driver, and it can add many levels on its own. Other common additions include the number of victims, whether the offense used “sophisticated means,” and the defendant’s role in the scheme.
In a drug case, the driving factor is usually the type and quantity of the controlled substance, with enhancements for things like weapons or the defendant’s role. The point is that two people convicted under the same statute can end up at very different offense levels depending on these facts.
Our managing partner, Russell Lorfing, describes the fraud guideline, USSG §2B1.1, as a menu of sentencing upgrades you never asked for. The real danger is stacking. A case that begins at a base offense level of 7 — relatively minor — can quietly climb into the mid-20s once loss amount, victim count, and enhancements are added on top of one another.
That is not a few extra months. As he puts it, it can be the difference between a halfway house and a high-security fence, which is why treating something as “just a simple fraud case” is a mistake.
The real danger is stacking. A case that starts at level 7 — relatively minor — can quietly climb into the mid-20s once loss amount, victim count, and enhancements are added on top of one another. That is not a rounding error. It is the difference between a halfway house and a high-security fence.
Step Three: Chapter 3 Adjustments
After the offense-specific factors, the guidelines apply general adjustments that can raise or lower the level.
Role in the offense. A leader or organizer of a scheme can receive an increase; a minor or minimal participant can receive a decrease.
Obstruction of justice. Lying under oath, destroying evidence, or interfering with the investigation can add levels.
Vulnerable victim or abuse of trust. Targeting a vulnerable victim or using a position of trust to commit the offense can increase the level.
Acceptance of responsibility. This is the most common reduction. A defendant who accepts responsibility — usually by pleading guilty in a timely way — can have the offense level reduced, which often makes a meaningful difference in the final range.
The result after all of this is the total offense level.
Step Four: Criminal History
The other axis of the calculation is your criminal history. Prior convictions are assigned points based on how serious they were and how recent, and the total points place you in one of six criminal history categories, from Category I (little or no history) to Category VI (extensive history).
A more serious criminal history moves you into a higher category, which increases the recommended sentence even when the offense level stays the same. Certain repeat classifications, such as the career offender designation, can raise the numbers dramatically.
Step Five: The Sentencing Table
The final guideline range comes from a table. One axis is the total offense level; the other is the criminal history category. Where they meet is a range expressed in months.
A higher offense level or a higher criminal history category moves you into a longer range. This is why shaving even a few levels off the offense level, or keeping a prior conviction from counting, can translate into years rather than months.
What the Defense Can Challenge at Each Step
The value of understanding the calculation is that nearly every input is contestable. A defense does not just argue for mercy at the end; it fights over the numbers that build the range.
The right guideline and base level. Which guideline applies, and therefore the base level, is sometimes disputable. Getting this right at the foundation affects everything above it.
The loss amount. In fraud cases, the government’s loss figure does the heavy lifting, and it is often built on assumptions that can be challenged point by point. Because reducing the loss can drop the offense level substantially, it is where much of the fight happens — covered in detail below.
Drug quantity. In drug cases, quantity drives the level. Counsel can contest how the quantity was calculated, what was reasonably attributable to the defendant, and whether estimates are reliable.
Enhancements. Sophisticated means, leadership role, number of victims, and obstruction are all enhancements the government reaches for, and each has legal requirements that can be contested on the facts.
Relevant conduct. This one surprises people, so it deserves its own section below.
Criminal history. How prior convictions are scored is not automatic. Counsel can argue that certain priors should not count, are too old, or were miscounted, which can move you into a lower category.
Acceptance of responsibility. Counsel works to secure this reduction and to protect it from being denied.
If you are facing sentencing, these are not abstract points — each one can be worth real time. Speak with our team about which parts of your calculation can be contested.
Loss Amount: Where Fraud Sentences Are Really Won
In a fraud case, the loss amount is the engine of the sentence. If loss drives the number, then reducing loss is how you cut the fuel line.
The government tends to present its loss figure as if it were handed down on stone tablets. It was not. It is usually a spreadsheet full of assumptions, and the work is poking holes in each one. To count as loss under §2B1.1, the harm has to be reasonably foreseeable, not purely speculative — and that is where the real fight happens.
Three questions do most of the work:
- Credits against loss. Did the defendant provide actual services, value, or collateral? That should come off the top.
- Double counting. Is the government counting the same dollar twice by attributing it to more than one victim?
- Speculative intended loss. Was that large number actually achievable, or just the biggest figure the government could reach for?
Because loss thresholds work like rungs on a ladder — the offense level jumps as the figure crosses set amounts — knocking even a modest sum off the calculation does not just save money. It can move you down a rung and give back years, not months.
Actual Loss vs. Intended Loss
One nuance drives a lot of loss litigation: the difference between what was actually lost and what the government says a person intended to take. Prosecutors often argue for the larger, intended figure because it gives them maximum leverage.
As an illustration, imagine someone attempted to move $5 million out of an account, but a technical failure — say the bank’s connection dropped — meant only $50 actually transferred. The government will typically argue the person should be sentenced as a $5 million offender, not a $50 one. That is exactly why loss becomes a “trial within the case,” and why failing to challenge the government’s math from the start lets prosecutors set the length of a sentence before the judge has really begun.
Sophisticated Means
“Sophisticated means” is one of the enhancements prosecutors reach for most, and the name oversells it. It does not require an Ocean’s Eleven scheme. Courts have applied it to conduct as ordinary as moving money between a main account and a shell account to mask where funds came from — in other words, you do not have to be a criminal mastermind to get hit with it.
There are three common ways to push back. First, the standard-business defense: the entities involved were legitimate LLCs formed for ordinary tax or business reasons. Second, the deconstruction argument: the conduct was not a layered conspiracy but simple, repetitive transfers. Third, proportionality: compared to a genuinely sophisticated operation, this was paperwork, not wizardry. If the conduct was routine, the enhancement should not apply.
Victim Enhancements
In a fraud case, the government does not just look at the money — it looks at the number of people affected. Under §2B1.1, the offense level increases by 2 if the offense involved 10 or more victims or was mass-marketed, and it climbs higher — by 4 or even 6 levels — as more victims suffer what the guidelines call “substantial financial hardship.”
The catch is that the definition of a victim can be broad, and people who were only lightly affected can still end up on the list. That makes the list worth going through with a fine-tooth comb, looking for duplicate entries, people who were already reimbursed, or “victims” whose loss actually came from something unrelated to the charged conduct. These numbers move judges, so they need to be accurate.
A Worked Example: How the Levels Stack Up
To see how the pieces combine, consider a hypothetical drawn from a common fraud type — a fraudulent Paycheck Protection Program loan. This example is illustrative only and does not reflect any actual client matter.
Imagine a $300,000 fraudulent PPP loan. The base offense level for the fraud is 7. The loss amount adds roughly 12 levels. Then come enhancements: using a fake identity might add 2 levels, and a sophisticated-means enhancement might add another 2. Just like that, the calculation reaches about level 23 — which, for someone with little criminal history, can translate into a multi-year prison range.
Now assume the defense goes to work. Securing acceptance of responsibility and successfully knocking out the sophisticated-means enhancement could bring the offense level down several rungs, potentially into the high teens. In prison terms, that is not a rounding error — it is a substantial difference in time. The lesson is simple: at sentencing, you do not just plead, you litigate the math.
At sentencing, you don't just plead. You litigate the math.
Outcomes always depend on the specific facts, and no result can be promised in advance.
Relevant Conduct: Why Your Sentence Can Reach Beyond the Charge
One of the least understood features of federal sentencing is relevant conduct. The guidelines allow a judge to consider not just the specific count you were convicted of, but a broader range of related conduct.
In practice, this means a sentence can be driven by conduct that was never charged, as long as the judge finds it happened by a lower standard of proof — a preponderance of the evidence, rather than the “beyond a reasonable doubt” standard required for a conviction. A fraud sentence, for example, can be based on the full scope of a scheme rather than only the counts in the plea. (A 2024 amendment to the guidelines now generally bars counting conduct a jury actually acquitted you of, though uncharged and dismissed conduct can still be considered.)
Because relevant conduct can enlarge a sentence dramatically, it is a major battleground. Counsel can contest what conduct is truly related, what is reasonably attributable to the defendant, and whether the government has actually proven it.
Departures and Variances: Two Ways Below the Range
Once the range is calculated, there are two distinct paths to a lower sentence, and the difference matters.
A departure is a move away from the range that the guidelines themselves authorize for specific reasons — for example, when the guideline range overstates the seriousness of the conduct, or on the government’s motion for substantial assistance.
A variance is different. Because the guidelines are advisory, a judge can impose a sentence outside the range based on the broader sentencing factors Congress set out. This is where much of the real advocacy happens.
The §3553(a) Factors
Federal law requires the judge to consider a set of factors, found in 18 U.S.C. § 3553(a), in deciding on a sentence that is “sufficient, but not greater than necessary.” These include the nature of the offense, the history and characteristics of the defendant, the need for the sentence to reflect the seriousness of the crime and provide deterrence, and the need to avoid unwarranted disparities.
This is where a defendant becomes a person rather than a point total. A strong sentencing presentation — showing who the defendant is, what led to the offense, what they have done since, and their prospects going forward — can persuade a judge to vary below the calculated range. That work is often the difference between a guideline sentence and something well below it.
Mandatory Minimums Cut Through the Whole Calculation
There is one situation where the guideline math takes a back seat. Some federal statutes carry mandatory minimum sentences set by Congress.
When a mandatory minimum applies, the judge generally cannot go below it, even if the calculated guideline range or the §3553(a) factors would support a lower sentence. Mandatory minimums appear most often in certain drug, firearm, and child-exploitation offenses.
There are limited paths around a mandatory minimum — most commonly a “safety valve” for certain low-level, nonviolent drug defendants who meet strict criteria, or a government motion recognizing substantial assistance. Whether either applies is a fact-specific question, and it is one of the most important early questions in any case carrying a mandatory minimum.
If your case may involve a mandatory minimum, understanding your options early is critical. Talk with our team about where your matter stands.
Why the Early Stages Affect the Sentence
Sentencing does not begin at the sentencing hearing. Decisions made much earlier — during the investigation, in interviews, in the plea negotiation — shape the numbers that show up later.
A statement to agents can support an obstruction enhancement. The way a plea is structured can affect the loss amount, the acceptance-of-responsibility reduction, and which conduct counts. This is why involving experienced counsel early, not just before the hearing, protects your sentencing position.
Why Clients Bring Us In for Federal Sentencing
Our firm’s managing partner, Russell Lorfing, is a former Assistant U.S. Attorney who prosecuted federal cases in Lubbock, Texas, and was designated a Cyber Hacking Intellectual Property Prosecutor by the U.S. Attorney’s Office in 2017. He has served as lead counsel in complex federal cases and has trained federal agents and prosecutors for the FBI, IRS, DEA, and DOJ. In 2024, he served as Co-Chair of the Federal Criminal Defense Committee for the Texas Criminal Defense Lawyers Association. He is admitted to practice in Texas, the District of Columbia, the Fifth Circuit Court of Appeals, the Northern, Western, and Southern Districts of Texas, and the U.S. Supreme Court.
Having calculated and argued these ranges from the prosecution side, he understands where the government’s numbers are soft — inflated loss figures, aggressive enhancements, and relevant-conduct theories that do not hold up. That is often where a sentence is won.
Founding partner Trey Keith has spent more than 20 years defending clients in state and federal matters, including financial crimes such as money laundering, with not-guilty verdicts in those cases. The Honorable E. Scott Frost (Ret.), Of Counsel to the firm, brings more than 30 years on the federal bench — the same vantage point from which sentences are actually imposed. Our broader team includes former federal prosecutors, former federal public defenders, former FBI agents, former IRS criminal investigators, and former general counsel to major family offices.
The guideline range is where sentencing starts, not where it has to end. To understand what the sentencing day itself looks like, see our page on what happens at a federal sentencing hearing. If your case involves specific conduct like securities, healthcare billing, or drug offenses, our pages on SEC Fraud Defense, Healthcare Fraud Defense, and Federal Drug Crime Defense explain how sentencing plays out in those areas.
Frequently Asked Questions
How is a federal sentence calculated?
The court finds a base offense level for the crime, adds levels for specific offense characteristics like loss amount or drug quantity, applies adjustments for things like role and acceptance of responsibility to reach a total offense level, then combines that with your criminal history category on the sentencing table to get a recommended range in months.
Does the judge have to follow the guideline range?
No. The guidelines are advisory. The judge must calculate and consider the range but can impose a sentence above or below it based on the sentencing factors in 18 U.S.C. § 3553(a).
What is the difference between a departure and a variance?
A departure is a move away from the range that the guidelines themselves authorize for specific reasons. A variance is a sentence outside the range based on the broader statutory sentencing factors, which is possible because the guidelines are advisory.
Can my sentence be lower than the guideline range?
Yes. Through a departure, a variance based on the §3553(a) factors, acceptance of responsibility, the safety valve in certain drug cases, or a government motion for substantial assistance, a sentence can fall below the calculated range — unless a mandatory minimum sets a floor.
What is acceptance of responsibility worth?
It is a reduction in the offense level, typically for defendants who plead guilty in a timely way. Because it lowers the offense level, it can move you into a shorter range on the sentencing table, and protecting this reduction is an important part of the defense.
What is relevant conduct?
Relevant conduct is a rule that lets the judge consider related conduct beyond the specific count of conviction — including uncharged conduct — when it is proven by a lower standard. A 2024 guideline amendment now generally excludes conduct a jury acquitted you of from that calculation. Relevant conduct can increase a sentence significantly, which is why it is heavily contested.
What are the §3553(a) factors?
They are the statutory factors a judge must weigh, including the nature of the offense, the defendant’s history and characteristics, the need for deterrence and to reflect the seriousness of the crime, and the need to avoid unwarranted disparities. They guide the judge toward a sentence that is sufficient but not greater than necessary.
Can the government’s loss amount be challenged?
Yes. Loss figures are often built on assumptions and can be contested — for overstatement, for failing to credit value actually provided, for double-counting, or for relying on speculative intended-loss numbers. Reducing the loss can lower the offense level and the resulting range.
If you are facing federal sentencing, the calculation is not final, and the earlier counsel gets involved, the more of it can be shaped. Schedule a consultation with our federal defense team to review your case.